ForeTiS.model._stat_model

Module Contents

Classes

StatModel

Parent class based on BaseModel for all models with a statsmodels-like API to share functionalities.
class ForeTiS.model._stat_model.StatModel(optuna_trial, datasets, featureset_name, optimize_featureset, current_model_name=None, target_column=None, pca_transform=None)

Bases: ForeTiS.model._base_model.BaseModel, abc.ABC

Parent class based on BaseModel for all models with a statsmodels-like API to share functionalities. See BaseModel for more information.

Parameters:

  • optuna_trial (optuna.trial.Trial) –

  • datasets (list) –

  • featureset_name (str) –

  • optimize_featureset (bool) –

  • current_model_name (str) –

  • target_column (str) –

  • pca_transform (bool) –

retrain(retrain)

Implementation of the retraining for models with statsmodels-like API. See BaseModel for more information.

Parameters:

retrain (pandas.DataFrame) –

update(update, period)

Update existing model due to new samples. See BaseModel for more information.

Parameters:

  • update (pandas.DataFrame) –

  • period (int) –

predict(X_in)

Implementation of a prediction based on input features for models with statsmodels-like API. See BaseModel for more information.

Parameters:

X_in (pandas.DataFrame) –

Return type:

numpy.array

train_val_loop(train, val)

Implementation of a train and validation loop for models with statsmodels-like API. See BaseModel for more information.

Parameters:

  • train (pandas.DataFrame) –

  • val (pandas.DataFrame) –

Return type:

numpy.array

get_transformed_set(df, target_column, transf, power_transformer, only_transform=False)

Function returning dataset with (log or power) transformed column

Parameters:

  • df (pandas.DataFrame) – dataset to transform

  • target_column (str) – column to transform

  • transf (str) – type of transformation

  • power_transformer (sklearn.preprocessing.PowerTransformer) – if power transforming was applied, the used transformer

  • only_transform – whether to only transform or not

Returns:

dataset with transformed column

Return type:

pandas.DataFrame

get_inverse_transformed_set(y, transf, power_transformer, is_conf=False)

Function returning inverse (log or power) transformed column

Parameters:

  • y (numpy.array) – array to be inverse transformed

  • power_transformer – if power transforming was applied, the used transformer

  • transf (str) – type of transformation

  • is_conf (bool) –

Returns:

transformed column

Return type:

numpy.array

static common_hyperparams()

Add hyperparameters that are common for PyTorch models. Do not need to be included in optimization for every child model. See BaseModel for more information.