pytabkit.models.sklearn package

Submodules

pytabkit.models.sklearn.default_params module

class pytabkit.models.sklearn.default_params.DefaultParams

Bases: object

CB_D = {'n_estimators': 1000}
CB_TD_CLASS = {'boosting_type': 'Plain', 'bootstrap_type': 'Bernoulli', 'early_stopping_rounds': 300, 'l2_leaf_reg': 1e-05, 'leaf_estimation_iterations': 1, 'lr': 0.08, 'max_bin': 254, 'max_depth': 7, 'n_estimators': 1000, 'one_hot_max_size': 15, 'random_strength': 0.8, 'subsample': 0.9}
CB_TD_REG = {'boosting_type': 'Plain', 'bootstrap_type': 'Bernoulli', 'early_stopping_rounds': 300, 'l2_leaf_reg': 1e-05, 'leaf_estimation_iterations': 20, 'lr': 0.09, 'max_bin': 254, 'max_depth': 9, 'n_estimators': 1000, 'one_hot_max_size': 20, 'random_strength': 0.0, 'subsample': 0.9}
FTT_D_CLASS = {'batch_size': 256, 'es_patience': 16, 'lr': 0.0001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 300, 'module_activation': 'reglu', 'module_attention_dropout': 0.2, 'module_d_ffn_factor': 1.3333333333333333, 'module_d_token': 192, 'module_ffn_dropout': 0.1, 'module_initialization': 'kaiming', 'module_kv_compression': None, 'module_kv_compression_sharing': None, 'module_n_heads': 8, 'module_n_layers': 3, 'module_prenormalization': True, 'module_residual_dropout': 0.0, 'module_token_bias': True, 'optimizer': 'adamw', 'optimizer_weight_decay': 1e-05, 'tfms': ['quantile_tabr'], 'use_checkpoints': True, 'verbose': 0}
FTT_D_REG = {'batch_size': 256, 'es_patience': 16, 'lr': 0.0001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 300, 'module_activation': 'reglu', 'module_attention_dropout': 0.2, 'module_d_ffn_factor': 1.3333333333333333, 'module_d_token': 192, 'module_ffn_dropout': 0.1, 'module_initialization': 'kaiming', 'module_kv_compression': None, 'module_kv_compression_sharing': None, 'module_n_heads': 8, 'module_n_layers': 3, 'module_prenormalization': True, 'module_residual_dropout': 0.0, 'module_token_bias': True, 'optimizer': 'adamw', 'optimizer_weight_decay': 1e-05, 'tfms': ['quantile_tabr'], 'transformed_target': True, 'use_checkpoints': True, 'verbose': 0}
LGBM_D = {'n_estimators': 100}
LGBM_TD_CLASS = {'bagging_freq': 1, 'colsample_bytree': 1.0, 'early_stopping_rounds': 300, 'lr': 0.04, 'max_bin': 255, 'min_data_in_leaf': 40, 'min_sum_hessian_in_leaf': 1e-07, 'n_estimators': 1000, 'num_leaves': 50, 'subsample': 0.75}
LGBM_TD_REG = {'bagging_freq': 1, 'colsample_bytree': 1.0, 'early_stopping_rounds': 300, 'lr': 0.05, 'max_bin': 255, 'min_data_in_leaf': 3, 'min_sum_hessian_in_leaf': 1e-07, 'n_estimators': 1000, 'num_leaves': 100, 'subsample': 0.7}
MLP_PLR_D_CLASS = {'batch_size': 128, 'es_patience': 20, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 1000, 'module_d_embedding': 8, 'module_d_first_layer': 128, 'module_d_last_layer': 128, 'module_d_layers': [128, 256, 128], 'module_dropout': 0.1, 'module_n_layers': 3, 'module_num_emb_dim': 24, 'module_num_emb_hidden_dim': 48, 'module_num_emb_lite': False, 'module_num_emb_sigma': 0.01, 'module_num_emb_type': 'plr', 'optimizer': 'adamw', 'tfms': ['quantile_tabr'], 'use_checkpoints': True, 'verbose': 0}
MLP_PLR_D_REG = {'batch_size': 128, 'es_patience': 20, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 1000, 'module_d_embedding': 8, 'module_d_first_layer': 128, 'module_d_last_layer': 128, 'module_d_layers': [128, 256, 128], 'module_dropout': 0.1, 'module_n_layers': 3, 'module_num_emb_dim': 24, 'module_num_emb_hidden_dim': 48, 'module_num_emb_lite': False, 'module_num_emb_sigma': 0.01, 'module_num_emb_type': 'plr', 'optimizer': 'adamw', 'tfms': ['quantile_tabr'], 'transformed_target': True, 'use_checkpoints': True, 'verbose': 0}
MLP_RTDL_D_CLASS_Grinsztajn = {'batch_size': 256, 'es_patience': 40, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 300, 'module_d_embedding': 128, 'module_d_first_layer': 128, 'module_d_last_layer': 128, 'module_d_layers': 256, 'module_dropout': 0.2, 'module_n_layers': 8, 'optimizer': 'adamw', 'tfms': ['quantile'], 'use_checkpoints': True, 'verbose': 0}
MLP_RTDL_D_CLASS_TabZilla = {'batch_size': 128, 'es_patience': 20, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 1000, 'module_d_embedding': 8, 'module_d_first_layer': 128, 'module_d_last_layer': 128, 'module_d_layers': [128, 256, 128], 'module_dropout': 0.1, 'module_n_layers': 3, 'optimizer': 'adamw', 'tfms': ['quantile_tabr'], 'use_checkpoints': True, 'verbose': 0}
MLP_RTDL_D_REG_Grinsztajn = {'batch_size': 256, 'es_patience': 40, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 300, 'module_d_embedding': 128, 'module_d_first_layer': 128, 'module_d_last_layer': 128, 'module_d_layers': 256, 'module_dropout': 0.2, 'module_n_layers': 8, 'optimizer': 'adamw', 'tfms': ['quantile'], 'transformed_target': True, 'use_checkpoints': True, 'verbose': 0}
MLP_RTDL_D_REG_TabZilla = {'batch_size': 128, 'es_patience': 20, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 1000, 'module_d_embedding': 8, 'module_d_first_layer': 128, 'module_d_last_layer': 128, 'module_d_layers': [128, 256, 128], 'module_dropout': 0.1, 'module_n_layers': 3, 'optimizer': 'adamw', 'tfms': ['quantile_tabr'], 'transformed_target': True, 'use_checkpoints': True, 'verbose': 0}
MLP_SKL_D = {'tfms': ['mean_center', 'l2_normalize', 'one_hot']}
RESNET_RTDL_D_CLASS_Grinsztajn = {'batch_size': 256, 'es_patience': 40, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 300, 'module_activation': 'reglu', 'module_d': 256, 'module_d_embedding': 128, 'module_d_hidden_factor': 2, 'module_hidden_dropout': 0.2, 'module_n_layers': 8, 'module_normalization': 'batchnorm', 'module_residual_dropout': 0.2, 'optimizer': 'adamw', 'optimizer_weight_decay': 1e-07, 'tfms': ['quantile'], 'use_checkpoints': True, 'verbose': 0}
RESNET_RTDL_D_CLASS_TabZilla = {'batch_size': 128, 'es_patience': 20, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 1000, 'module_activation': 'relu', 'module_d': 128, 'module_d_embedding': 8, 'module_d_hidden_factor': 2, 'module_hidden_dropout': 0.25, 'module_n_layers': 2, 'module_normalization': 'batchnorm', 'module_residual_dropout': 0.1, 'optimizer': 'adamw', 'optimizer_weight_decay': 0.01, 'tfms': ['quantile_tabr'], 'use_checkpoints': True, 'verbose': 0}
RESNET_RTDL_D_REG_Grinsztajn = {'batch_size': 256, 'es_patience': 40, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 300, 'module_activation': 'reglu', 'module_d': 256, 'module_d_embedding': 128, 'module_d_hidden_factor': 2, 'module_hidden_dropout': 0.2, 'module_n_layers': 8, 'module_normalization': 'batchnorm', 'module_residual_dropout': 0.2, 'optimizer': 'adamw', 'optimizer_weight_decay': 1e-07, 'tfms': ['quantile'], 'transformed_target': True, 'use_checkpoints': True, 'verbose': 0}
RESNET_RTDL_D_REG_TabZilla = {'batch_size': 128, 'es_patience': 20, 'lr': 0.001, 'lr_patience': 30, 'lr_scheduler': False, 'max_epochs': 1000, 'module_activation': 'relu', 'module_d': 128, 'module_d_embedding': 8, 'module_d_hidden_factor': 2, 'module_hidden_dropout': 0.25, 'module_n_layers': 2, 'module_normalization': 'batchnorm', 'module_residual_dropout': 0.1, 'optimizer': 'adamw', 'optimizer_weight_decay': 0.01, 'tfms': ['quantile_tabr'], 'transformed_target': True, 'use_checkpoints': True, 'verbose': 0}
RF_SKL_D = {'permute_ordinal_encoding': True, 'tfms': ['ordinal_encoding']}
RealMLP_TD_CLASS = {'act': 'selu', 'act_lr_factor': 0.1, 'add_front_scale': True, 'bias_init_mode': 'he+5', 'bias_lr_factor': 0.1, 'bias_wd_factor': 0.0, 'block_str': 'w-b-a-d', 'embedding_size': 8, 'hidden_sizes': [256, 256, 256], 'lr': 0.04, 'lr_sched': 'coslog4', 'ls_eps': 0.1, 'max_one_hot_cat_size': 9, 'n_epochs': 256, 'num_emb_type': 'pbld', 'opt': 'adam', 'p_drop': 0.15, 'p_drop_sched': 'flat_cos', 'plr_hidden_1': 16, 'plr_hidden_2': 4, 'plr_lr_factor': 0.1, 'plr_sigma': 0.1, 'scale_lr_factor': 6.0, 'sq_mom': 0.95, 'tfms': ['one_hot', 'median_center', 'robust_scale', 'smooth_clip', 'embedding'], 'use_ls': True, 'use_parametric_act': True, 'wd': 0.02, 'wd_sched': 'flat_cos', 'weight_init_mode': 'std', 'weight_param': 'ntk'}
RealMLP_TD_REG = {'act': 'mish', 'act_lr_factor': 0.1, 'add_front_scale': True, 'bias_init_mode': 'he+5', 'bias_lr_factor': 0.1, 'bias_wd_factor': 0.0, 'block_str': 'w-b-a-d', 'clamp_output': True, 'embedding_size': 8, 'hidden_sizes': [256, 256, 256], 'lr': 0.2, 'lr_sched': 'coslog4', 'max_one_hot_cat_size': 9, 'n_epochs': 256, 'normalize_output': True, 'num_emb_type': 'pbld', 'opt': 'adam', 'p_drop': 0.15, 'p_drop_sched': 'flat_cos', 'plr_hidden_1': 16, 'plr_hidden_2': 4, 'plr_lr_factor': 0.1, 'plr_sigma': 0.1, 'scale_lr_factor': 6.0, 'sq_mom': 0.95, 'tfms': ['one_hot', 'median_center', 'robust_scale', 'smooth_clip', 'embedding'], 'use_parametric_act': True, 'wd': 0.02, 'wd_sched': 'flat_cos', 'weight_init_mode': 'std', 'weight_param': 'ntk'}
RealMLP_TD_S_CLASS = {'act': 'selu', 'add_front_scale': True, 'bias_init_mode': 'normal', 'bias_lr_factor': 0.1, 'block_str': 'w-b-a', 'hidden_sizes': [256, 256, 256], 'last_layer_config': {'bias_init_mode': 'zeros', 'weight_init_mode': 'zeros'}, 'lr': 0.04, 'lr_sched': 'coslog4', 'ls_eps': 0.1, 'n_epochs': 256, 'opt': 'adam', 'scale_lr_factor': 6.0, 'sq_mom': 0.95, 'tfms': ['one_hot', 'median_center', 'robust_scale', 'smooth_clip'], 'use_ls': True, 'weight_init_mode': 'normal', 'weight_param': 'ntk'}
RealMLP_TD_S_REG = {'act': 'mish', 'add_front_scale': True, 'bias_init_mode': 'normal', 'bias_lr_factor': 0.1, 'block_str': 'w-b-a', 'hidden_sizes': [256, 256, 256], 'last_layer_config': {'bias_init_mode': 'zeros', 'weight_init_mode': 'zeros'}, 'lr': 0.07, 'lr_sched': 'coslog4', 'n_epochs': 256, 'normalize_output': True, 'opt': 'adam', 'scale_lr_factor': 6.0, 'sq_mom': 0.95, 'tfms': ['one_hot', 'median_center', 'robust_scale', 'smooth_clip'], 'weight_init_mode': 'normal', 'weight_param': 'ntk'}
RealTABR_D_CLASS = {'activation': 'ReLU', 'add_scaling_layer': True, 'batch_size': 'auto', 'context_dropout': 0.38920071545944357, 'context_size': 96, 'd_main': 265, 'd_multiplier': 2.0, 'dropout0': 0.38852797479169876, 'dropout1': 0.0, 'encoder_n_blocks': 0, 'eval_batch_size': 4096, 'freeze_contexts_after_n_epochs': None, 'ls_eps': 0.1, 'mixer_normalization': 'auto', 'n_epochs': 100000, 'normalization': 'LayerNorm', 'num_embeddings': {'d_embedding': 4, 'frequency_scale': 0.1, 'n_frequencies': 8, 'type': 'PBLDEmbeddings'}, 'optimizer': {'betas': (0.9, 0.95), 'lr': 0.0003121273641315169, 'type': 'AdamW', 'weight_decay': 1.2260352006404615e-06}, 'patience': 16, 'predictor_n_blocks': 1, 'scale_lr_factor': 96, 'tfms': ['median_center', 'robust_scale', 'smooth_clip']}
RealTABR_D_REG = {'activation': 'ReLU', 'add_scaling_layer': True, 'batch_size': 'auto', 'context_dropout': 0.38920071545944357, 'context_size': 96, 'd_main': 265, 'd_multiplier': 2.0, 'dropout0': 0.38852797479169876, 'dropout1': 0.0, 'encoder_n_blocks': 0, 'eval_batch_size': 4096, 'freeze_contexts_after_n_epochs': None, 'ls_eps': 0.1, 'mixer_normalization': 'auto', 'n_epochs': 100000, 'normalization': 'LayerNorm', 'num_embeddings': {'d_embedding': 4, 'frequency_scale': 0.1, 'n_frequencies': 8, 'type': 'PBLDEmbeddings'}, 'optimizer': {'betas': (0.9, 0.95), 'lr': 0.0003121273641315169, 'type': 'AdamW', 'weight_decay': 1.2260352006404615e-06}, 'patience': 16, 'predictor_n_blocks': 1, 'scale_lr_factor': 96, 'tfms': ['median_center', 'robust_scale', 'smooth_clip'], 'transformed_target': True}
TABM_D_CLASS = {'allow_amp': False, 'arch_type': 'tabm', 'batch_size': 256, 'compile_model': False, 'd_block': 512, 'd_embedding': 16, 'dropout': 0.1, 'gradient_clipping_norm': None, 'lr': 0.002, 'n_blocks': 'auto', 'n_epochs': 1000000000, 'num_emb_n_bins': 48, 'num_emb_type': 'none', 'patience': 16, 'tabm_k': 32, 'tfms': ['quantile_tabr'], 'weight_decay': 0.0}
TABM_D_REG = {'allow_amp': False, 'arch_type': 'tabm', 'batch_size': 256, 'compile_model': False, 'd_block': 512, 'd_embedding': 16, 'dropout': 0.1, 'gradient_clipping_norm': None, 'lr': 0.002, 'n_blocks': 'auto', 'n_epochs': 1000000000, 'num_emb_n_bins': 48, 'num_emb_type': 'none', 'patience': 16, 'tabm_k': 32, 'tfms': ['quantile_tabr'], 'weight_decay': 0.0}
TABR_S_D_CLASS = {'activation': 'ReLU', 'batch_size': 'auto', 'context_dropout': 0.38920071545944357, 'context_size': 96, 'd_main': 265, 'd_multiplier': 2.0, 'dropout0': 0.38852797479169876, 'dropout1': 0.0, 'encoder_n_blocks': 0, 'eval_batch_size': 4096, 'freeze_contexts_after_n_epochs': None, 'mixer_normalization': 'auto', 'n_epochs': 100000, 'normalization': 'LayerNorm', 'num_embeddings': None, 'optimizer': {'lr': 0.0003121273641315169, 'type': 'AdamW', 'weight_decay': 1.2260352006404615e-06}, 'patience': 16, 'predictor_n_blocks': 1, 'tfms': ['quantile_tabr']}
TABR_S_D_CLASS_FREEZE = {'activation': 'ReLU', 'batch_size': 'auto', 'context_dropout': 0.38920071545944357, 'context_size': 96, 'd_main': 265, 'd_multiplier': 2.0, 'dropout0': 0.38852797479169876, 'dropout1': 0.0, 'encoder_n_blocks': 0, 'eval_batch_size': 4096, 'freeze_contexts_after_n_epochs': 4, 'mixer_normalization': 'auto', 'n_epochs': 100000, 'normalization': 'LayerNorm', 'num_embeddings': None, 'optimizer': {'lr': 0.0003121273641315169, 'type': 'AdamW', 'weight_decay': 1.2260352006404615e-06}, 'patience': 16, 'predictor_n_blocks': 1, 'tfms': ['quantile_tabr']}
TABR_S_D_REG = {'activation': 'ReLU', 'batch_size': 'auto', 'context_dropout': 0.38920071545944357, 'context_size': 96, 'd_main': 265, 'd_multiplier': 2.0, 'dropout0': 0.38852797479169876, 'dropout1': 0.0, 'encoder_n_blocks': 0, 'eval_batch_size': 4096, 'freeze_contexts_after_n_epochs': None, 'mixer_normalization': 'auto', 'n_epochs': 100000, 'normalization': 'LayerNorm', 'num_embeddings': None, 'optimizer': {'lr': 0.0003121273641315169, 'type': 'AdamW', 'weight_decay': 1.2260352006404615e-06}, 'patience': 16, 'predictor_n_blocks': 1, 'tfms': ['quantile_tabr'], 'transformed_target': True}
TABR_S_D_REG_FREEZE = {'activation': 'ReLU', 'batch_size': 'auto', 'context_dropout': 0.38920071545944357, 'context_size': 96, 'd_main': 265, 'd_multiplier': 2.0, 'dropout0': 0.38852797479169876, 'dropout1': 0.0, 'encoder_n_blocks': 0, 'eval_batch_size': 4096, 'freeze_contexts_after_n_epochs': 4, 'mixer_normalization': 'auto', 'n_epochs': 100000, 'normalization': 'LayerNorm', 'num_embeddings': None, 'optimizer': {'lr': 0.0003121273641315169, 'type': 'AdamW', 'weight_decay': 1.2260352006404615e-06}, 'patience': 16, 'predictor_n_blocks': 1, 'tfms': ['quantile_tabr'], 'transformed_target': True}
VANILLA_MLP_CLASS = {'act': 'relu', 'batch_size': 256, 'bias_init_mode': 'pytorch-default', 'block_str': 'w-b-a-d', 'hidden_sizes': [256, 256, 256], 'lr': 0.001, 'lr_sched': 'constant', 'max_n_vectorized': 1, 'n_epochs': 256, 'opt': 'adam', 'p_drop': 0.0, 'tfms': ['quantile', 'one_hot'], 'use_last_best_epoch': False, 'wd': 0.0, 'weight_init_gain': np.float64(0.5773502691896258), 'weight_init_mode': 'uniform', 'weight_param': 'standard'}
VANILLA_MLP_REG = {'act': 'relu', 'batch_size': 256, 'bias_init_mode': 'pytorch-default', 'block_str': 'w-b-a-d', 'hidden_sizes': [256, 256, 256], 'lr': 0.001, 'lr_sched': 'constant', 'max_n_vectorized': 1, 'n_epochs': 256, 'normalize_output': True, 'opt': 'adam', 'p_drop': 0.0, 'tfms': ['quantile', 'one_hot'], 'use_last_best_epoch': False, 'wd': 0.0, 'weight_init_gain': np.float64(0.5773502691896258), 'weight_init_mode': 'uniform', 'weight_param': 'standard'}
XGB_D = {'n_estimators': 100, 'tree_method': 'hist'}
XGB_PBB_CLASS = {'colsample_bylevel': 0.585, 'colsample_bytree': 0.752, 'lr': 0.018, 'max_depth': 13, 'max_n_threads': 64, 'max_one_hot_cat_size': 20, 'min_child_weight': 2.06, 'n_estimators': 4168, 'reg_alpha': 1.113, 'reg_lambda': 0.982, 'subsample': 0.839, 'tfms': ['one_hot'], 'tree_method': 'hist'}
XGB_TD_CLASS = {'colsample_bylevel': 0.9, 'early_stopping_rounds': 300, 'lr': 0.08, 'max_bin': 256, 'max_depth': 6, 'min_child_weight': 5e-06, 'n_estimators': 1000, 'reg_lambda': 0.0, 'subsample': 0.65, 'tree_method': 'hist'}
XGB_TD_REG = {'early_stopping_rounds': 300, 'lr': 0.05, 'max_bin': 256, 'max_depth': 9, 'min_child_weight': 2.0, 'n_estimators': 1000, 'reg_lambda': 0.0, 'subsample': 0.7, 'tree_method': 'hist'}
XRFM_D_CLASS = {'M_batch_size': 8000, 'bandwidth': 10.0, 'bandwidth_mode': 'constant', 'classification_mode': 'prevalence', 'diag': True, 'early_stop_multiplier': 1.1, 'early_stop_rfm': True, 'exponent': 1.0, 'iters': 5, 'kernel_type': 'l2', 'max_leaf_samples': 60000, 'p_interp': 1.0, 'reg': 0.001}
XRFM_D_REG = {'M_batch_size': 8000, 'bandwidth': 10.0, 'bandwidth_mode': 'constant', 'classification_mode': 'prevalence', 'diag': True, 'early_stop_multiplier': 1.1, 'early_stop_rfm': True, 'exponent': 1.0, 'iters': 5, 'kernel_type': 'l2', 'max_leaf_samples': 60000, 'p_interp': 1.0, 'reg': 0.001}

pytabkit.models.sklearn.sklearn_base module

class pytabkit.models.sklearn.sklearn_base.AlgInterfaceClassifier

Bases: ClassifierMixin, AlgInterfaceEstimator

predict(X)

Predict labels.

Parameters

Xarray-like, shape (n_samples, n_features)

The input samples.

Returns

yndarray, shape (n_samples,)

The label for each sample is the label of the closest sample seen during fit.

predict_ensemble(X)
predict_proba(X)
Return type:

ndarray

predict_proba_ensemble(X)
Return type:

ndarray

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (AlgInterfaceClassifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

AlgInterfaceClassifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

AlgInterfaceClassifier

class pytabkit.models.sklearn.sklearn_base.AlgInterfaceEstimator

Bases: BaseEstimator

Base class for wrapping AlgInterface subclasses with a scikit-learn compatible interface.

fit(X, y, X_val=None, y_val=None, val_idxs=None, cat_indicator=None, cat_col_names=None, time_to_fit_in_seconds=None)

Fit the estimator.

Parameters:

  • X – Inputs (covariates). pandas DataFrame, numpy array, or similar array-like.

  • y – Labels (targets, variates). pandas DataFrame/Series, numpy array, or similar array-like.

  • X_val (Optional) – Inputs for validation set. Can only be used if n_cv is not set to a value other than 1, and if val_idxs is not used. If X_val is used, X will be used for the training set only, instead of getting validation data from X.

  • y_val (Optional) – Labels for the validation set.

  • val_idxs (ndarray | None) – Indices of validation set elements within X and y (optional). Can be an array of shape (n_val_samples,) or (n_val_splits,n_val_samples_per_split). In the latter case, the results of the models on the validation splits will be ensembled.

  • cat_indicator (List[bool] | ndarray | None) – Which features/columns are categorical, specified as a list or array of booleans. If this is not specified, all columns with category/string/object dtypes are interpreted as categorical and all others as numerical.

  • cat_col_names (List[str] | None) – List of column names that should be treated as categorical (if X is a pd.DataFrame). Can be specified instead of cat_indicator.

  • time_to_fit_in_seconds (int | None) – Time limit in seconds for fitting. Currently only implemented for RealMLP (default=None). If None, no time limit will be applied.

Returns:

Returns self.

Return type:

BaseEstimator

get_config()

Augments the result from self.get_params() with the parameters from self._get_default_params(). Uses _preprocess_config_key() to change the names from self.get_params() if implemented. Default parameters are used if the value in get_params() is either None or not present. :return: Dictionary of parameters augmented with default parameters.

Return type:

Dict[str, Any]

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (AlgInterfaceEstimator)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

AlgInterfaceEstimator

to(device)

Move the model (only implemented for RealMLP at the moment) to the specified device. :param device: PyTorch-compatible device name.

Parameters:

device (str)

Return type:

None

class pytabkit.models.sklearn.sklearn_base.AlgInterfaceRegressor

Bases: RegressorMixin, AlgInterfaceEstimator

predict(X)
predict_ensemble(X)
set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (AlgInterfaceRegressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

AlgInterfaceRegressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

AlgInterfaceRegressor

pytabkit.models.sklearn.sklearn_base.check_X_y_wrapper(*args, **kwargs)
pytabkit.models.sklearn.sklearn_base.check_array_wrapper(*args, **kwargs)
pytabkit.models.sklearn.sklearn_base.concat_arrays(x1, x2)
Return type:

Any

pytabkit.models.sklearn.sklearn_base.to_df(x)
Return type:

DataFrame

pytabkit.models.sklearn.sklearn_base.to_normal_type(x)
Return type:

Any

pytabkit.models.sklearn.sklearn_interfaces module

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_D_Classifier

Bases: CatBoostConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_D_Regressor

Bases: CatBoostConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_HPO_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_HPO_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_HPO_TPE_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_HPO_TPE_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_HPO_TPE_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_HPO_TPE_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_HPO_TPE_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_HPO_TPE_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_HPO_TPE_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_HPO_TPE_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_TD_Classifier

Bases: CatBoostConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_TD_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_TD_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_TD_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.CatBoost_TD_Regressor

Bases: CatBoostConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (CatBoost_TD_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

CatBoost_TD_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

CatBoost_TD_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.Ensemble_HPO_Classifier

Bases: EnsembleHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Ensemble_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Ensemble_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Ensemble_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.Ensemble_HPO_Regressor

Bases: EnsembleHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Ensemble_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Ensemble_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Ensemble_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.Ensemble_TD_Classifier

Bases: AlgInterfaceClassifier

__init__(device=None, random_state=None, n_cv=1, n_refit=0, n_repeats=1, val_fraction=0.2, n_threads=None, tmp_folder=None, verbosity=0, val_metric_name=None, use_ls=None, calibration_method=None)
Parameters:

  • device (str | None)

  • random_state (int | RandomState | None)

  • n_cv (int)

  • n_refit (int)

  • n_repeats (int)

  • val_fraction (float)

  • n_threads (int | None)

  • tmp_folder (str | Path | None)

  • verbosity (int)

  • val_metric_name (str | None)

  • use_ls (bool | None)

  • calibration_method (str | None)

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Ensemble_TD_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Ensemble_TD_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Ensemble_TD_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.Ensemble_TD_Regressor

Bases: AlgInterfaceRegressor

__init__(device=None, random_state=None, n_cv=1, n_refit=0, n_repeats=1, val_fraction=0.2, n_threads=None, tmp_folder=None, verbosity=0, val_metric_name=None)
Parameters:

  • device (str | None)

  • random_state (int | RandomState | None)

  • n_cv (int)

  • n_refit (int)

  • n_repeats (int)

  • val_fraction (float)

  • n_threads (int | None)

  • tmp_folder (str | Path | None)

  • verbosity (int)

  • val_metric_name (str | None)

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Ensemble_TD_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Ensemble_TD_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Ensemble_TD_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.FTT_D_Classifier

Bases: FTTransformerConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (FTT_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

FTT_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

FTT_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.FTT_D_Regressor

Bases: FTTransformerConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (FTT_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

FTT_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

FTT_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.FTT_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (FTT_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

FTT_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

FTT_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.FTT_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (FTT_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

FTT_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

FTT_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_D_Classifier

Bases: LGBMConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_D_Regressor

Bases: LGBMConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_HPO_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_HPO_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_HPO_TPE_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_HPO_TPE_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_HPO_TPE_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_HPO_TPE_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_HPO_TPE_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_HPO_TPE_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_HPO_TPE_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_HPO_TPE_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_TD_Classifier

Bases: LGBMConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_TD_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_TD_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_TD_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.LGBM_TD_Regressor

Bases: LGBMConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (LGBM_TD_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

LGBM_TD_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

LGBM_TD_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.MLP_PLR_D_Classifier

Bases: RTDL_MLPConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_PLR_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_PLR_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_PLR_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.MLP_PLR_D_Regressor

Bases: RTDL_MLPConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_PLR_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_PLR_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_PLR_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.MLP_PLR_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_PLR_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_PLR_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_PLR_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.MLP_PLR_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_PLR_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_PLR_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_PLR_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.MLP_RTDL_D_Classifier

Bases: RTDL_MLPConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_RTDL_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_RTDL_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_RTDL_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.MLP_RTDL_D_Regressor

Bases: RTDL_MLPConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_RTDL_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_RTDL_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_RTDL_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.MLP_RTDL_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_RTDL_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_RTDL_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_RTDL_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.MLP_RTDL_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_RTDL_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_RTDL_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_RTDL_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.MLP_SKL_D_Classifier

Bases: MLPSKLConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_SKL_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_SKL_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_SKL_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.MLP_SKL_D_Regressor

Bases: MLPSKLConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (MLP_SKL_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

MLP_SKL_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

MLP_SKL_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.RF_HPO_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RF_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RF_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RF_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.RF_HPO_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RF_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RF_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RF_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.RF_SKL_D_Classifier

Bases: RFConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RF_SKL_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RF_SKL_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RF_SKL_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.RF_SKL_D_Regressor

Bases: RFConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RF_SKL_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RF_SKL_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RF_SKL_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.RealMLP_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealMLP_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealMLP_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealMLP_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.RealMLP_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealMLP_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealMLP_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealMLP_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.RealMLP_TD_Classifier

Bases: RealMLPConstructorMixin, AlgInterfaceClassifier

MLP-TD classifier. For constructor parameters, see MLPConstructorMixin.

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealMLP_TD_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealMLP_TD_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealMLP_TD_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.RealMLP_TD_Regressor

Bases: RealMLPConstructorMixin, AlgInterfaceRegressor

MLP-TD regressor. For constructor parameters, see MLPConstructorMixin.

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealMLP_TD_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealMLP_TD_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealMLP_TD_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.RealMLP_TD_S_Classifier

Bases: RealMLPConstructorMixin, AlgInterfaceClassifier

MLP-TD-S classifier. For constructor parameters, see MLPConstructorMixin.

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealMLP_TD_S_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealMLP_TD_S_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealMLP_TD_S_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.RealMLP_TD_S_Regressor

Bases: RealMLPConstructorMixin, AlgInterfaceRegressor

MLP-TD-S regressor. For constructor parameters, see MLPConstructorMixin.

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealMLP_TD_S_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealMLP_TD_S_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealMLP_TD_S_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.RealTabR_D_Classifier

Bases: TabrConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealTabR_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealTabR_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealTabR_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.RealTabR_D_Regressor

Bases: TabrConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (RealTabR_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

RealTabR_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

RealTabR_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.Resnet_RTDL_D_Classifier

Bases: ResnetConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Resnet_RTDL_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Resnet_RTDL_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Resnet_RTDL_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.Resnet_RTDL_D_Regressor

Bases: ResnetConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Resnet_RTDL_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Resnet_RTDL_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Resnet_RTDL_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.Resnet_RTDL_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Resnet_RTDL_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Resnet_RTDL_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Resnet_RTDL_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.Resnet_RTDL_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (Resnet_RTDL_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

Resnet_RTDL_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

Resnet_RTDL_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.TabM_D_Classifier

Bases: TabMConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabM_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabM_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabM_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.TabM_D_Regressor

Bases: TabMConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabM_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabM_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabM_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.TabM_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

HPO spaces (‘default’, ‘tabarena’) use TabM-mini with numerical embeddings

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabM_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabM_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabM_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.TabM_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

HPO spaces (‘default’, ‘tabarena’) use TabM-mini with numerical embeddings

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabM_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabM_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabM_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.TabR_HPO_Classifier

Bases: RealMLPHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabR_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabR_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabR_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.TabR_HPO_Regressor

Bases: RealMLPHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabR_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabR_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabR_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.TabR_S_D_Classifier

Bases: TabrConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabR_S_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabR_S_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabR_S_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.TabR_S_D_Regressor

Bases: TabrConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (TabR_S_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

TabR_S_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

TabR_S_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.XGB_D_Classifier

Bases: XGBConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XGB_D_Regressor

Bases: XGBConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.XGB_HPO_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XGB_HPO_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_HPO_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.XGB_HPO_TPE_Classifier

Bases: GBDTHPOConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_HPO_TPE_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_HPO_TPE_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_HPO_TPE_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XGB_HPO_TPE_Regressor

Bases: GBDTHPOConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_HPO_TPE_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_HPO_TPE_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_HPO_TPE_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.XGB_PBB_D_Classifier

Bases: XGBConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_PBB_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_PBB_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_PBB_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XGB_TD_Classifier

Bases: XGBConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_TD_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_TD_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_TD_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XGB_TD_Regressor

Bases: XGBConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XGB_TD_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XGB_TD_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XGB_TD_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.XRFM_D_Classifier

Bases: XRFMConstructorMixin, AlgInterfaceClassifier

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XRFM_D_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XRFM_D_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XRFM_D_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XRFM_D_Regressor

Bases: XRFMConstructorMixin, AlgInterfaceRegressor

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XRFM_D_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XRFM_D_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XRFM_D_Regressor

class pytabkit.models.sklearn.sklearn_interfaces.XRFM_HPO_Classifier

Bases: XRFMHPOConstructorMixin, AlgInterfaceClassifier

HPO spaces (‘default’) use xRFM

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XRFM_HPO_Classifier)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XRFM_HPO_Classifier

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XRFM_HPO_Classifier

class pytabkit.models.sklearn.sklearn_interfaces.XRFM_HPO_Regressor

Bases: XRFMHPOConstructorMixin, AlgInterfaceRegressor

HPO spaces (‘default’, ‘tabarena’) use TabM-mini with numerical embeddings

set_fit_request(*, X_val='$UNCHANGED$', cat_col_names='$UNCHANGED$', cat_indicator='$UNCHANGED$', time_to_fit_in_seconds='$UNCHANGED$', val_idxs='$UNCHANGED$', y_val='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to fit.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

X_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for X_val parameter in fit.

cat_col_namesstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_col_names parameter in fit.

cat_indicatorstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for cat_indicator parameter in fit.

time_to_fit_in_secondsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for time_to_fit_in_seconds parameter in fit.

val_idxsstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for val_idxs parameter in fit.

y_valstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for y_val parameter in fit.

Returns

selfobject

The updated object.

Parameters:

  • self (XRFM_HPO_Regressor)

  • X_val (bool | None | str)

  • cat_col_names (bool | None | str)

  • cat_indicator (bool | None | str)

  • time_to_fit_in_seconds (bool | None | str)

  • val_idxs (bool | None | str)

  • y_val (bool | None | str)

Return type:

XRFM_HPO_Regressor

set_score_request(*, sample_weight='$UNCHANGED$')

Configure whether metadata should be requested to be passed to the score method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

  • True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.

  • False: metadata is not requested and the meta-estimator will not pass it to score.

  • None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.

  • str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED

Metadata routing for sample_weight parameter in score.

Returns

selfobject

The updated object.

Parameters:
Return type:

XRFM_HPO_Regressor

Module contents