Skip to content

ForecasterAutoregDirect

ForecasterAutoregDirect(regressor, steps, lags, transformer_y=None, transformer_exog=None, weight_func=None, fit_kwargs=None, n_jobs='auto', forecaster_id=None)

Bases: ForecasterBase

This class turns any regressor compatible with the scikit-learn API into a autoregressive direct multi-step forecaster. A separate model is created for each forecast time step. See documentation for more details.

Parameters:

Name Type Description Default
regressor regressor or pipeline compatible with the scikit-learn API

An instance of a regressor or pipeline compatible with the scikit-learn API.

required
steps int

Maximum number of future steps the forecaster will predict when using method predict(). Since a different model is created for each step, this value should be defined before training.

required
lags int, list, numpy ndarray, range

Lags used as predictors. Index starts at 1, so lag 1 is equal to t-1.

  • int: include lags from 1 to lags (included).
  • list, 1d numpy ndarray or range: include only lags present in lags, all elements must be int.
required
transformer_y object transformer (preprocessor)

An instance of a transformer (preprocessor) compatible with the scikit-learn preprocessing API with methods: fit, transform, fit_transform and inverse_transform. ColumnTransformers are not allowed since they do not have inverse_transform method. The transformation is applied to y before training the forecaster.

`None`
transformer_exog object transformer (preprocessor)

An instance of a transformer (preprocessor) compatible with the scikit-learn preprocessing API. The transformation is applied to exog before training the forecaster. inverse_transform is not available when using ColumnTransformers.

`None`
weight_func Callable

Function that defines the individual weights for each sample based on the index. For example, a function that assigns a lower weight to certain dates. Ignored if regressor does not have the argument sample_weight in its fit method. The resulting sample_weight cannot have negative values.

`None`
fit_kwargs dict

Additional arguments to be passed to the fit method of the regressor.

`None`
n_jobs (int, auto)

The number of jobs to run in parallel. If -1, then the number of jobs is set to the number of cores. If 'auto', n_jobs is set using the function skforecast.utils.select_n_jobs_fit_forecaster. New in version 0.9.0

`'auto'`
forecaster_id (str, int)

Name used as an identifier of the forecaster.

`None`

Attributes:

Name Type Description
regressor regressor or pipeline compatible with the scikit-learn API

An instance of a regressor or pipeline compatible with the scikit-learn API. An instance of this regressor is trained for each step. All of them are stored in self.regressors_.

regressors_ dict

Dictionary with regressors trained for each step. They are initialized as a copy of regressor.

steps int

Number of future steps the forecaster will predict when using method predict(). Since a different model is created for each step, this value should be defined before training.

lags numpy ndarray

Lags used as predictors.

transformer_y object transformer (preprocessor)

An instance of a transformer (preprocessor) compatible with the scikit-learn preprocessing API with methods: fit, transform, fit_transform and inverse_transform. ColumnTransformers are not allowed since they do not have inverse_transform method. The transformation is applied to y before training the forecaster.

transformer_exog object transformer (preprocessor)

An instance of a transformer (preprocessor) compatible with the scikit-learn preprocessing API. The transformation is applied to exog before training the forecaster. inverse_transform is not available when using ColumnTransformers.

weight_func Callable

Function that defines the individual weights for each sample based on the index. For example, a function that assigns a lower weight to certain dates. Ignored if regressor does not have the argument sample_weight in its fit method. The resulting sample_weight cannot have negative values.

source_code_weight_func str

Source code of the custom function used to create weights.

max_lag int

Maximum value of lag included in lags.

window_size int

Size of the window needed to create the predictors. It is equal to max_lag.

window_size_diff int

This attribute has the same value as window_size as this Forecaster doesn't support differentiation.

last_window pandas DataFrame

Last window the forecaster has seen during training. It stores the values needed to predict steps immediately after the training data.

index_type type

Type of index of the input used in training.

index_freq str

Frequency of Index of the input used in training.

training_range pandas Index

First and last values of index of the data used during training.

included_exog bool

If the forecaster has been trained using exogenous variable/s.

exog_type type

Type of exogenous variable/s used in training.

exog_dtypes dict

Type of each exogenous variable/s used in training. If transformer_exog is used, the dtypes are calculated after the transformation.

exog_col_names list

Names of the exogenous variables used during training.

X_train_col_names list

Names of columns of the matrix created internally for training.

fit_kwargs dict

Additional arguments to be passed to the fit method of the regressor.

in_sample_residuals dict

Residuals of the models when predicting training data. Only stored up to 1000 values per model in the form {step: residuals}. If transformer_y is not None, residuals are stored in the transformed scale.

out_sample_residuals dict

Residuals of the models when predicting non training data. Only stored up to 1000 values per model in the form {step: residuals}. If transformer_y is not None, residuals are assumed to be in the transformed scale. Use set_out_sample_residuals() method to set values.

fitted bool

Tag to identify if the regressor has been fitted (trained).

creation_date str

Date of creation.

fit_date str

Date of last fit.

skforecast_version str

Version of skforecast library used to create the forecaster.

python_version str

Version of python used to create the forecaster.

n_jobs int, 'auto', default `'auto'`

The number of jobs to run in parallel. If -1, then the number of jobs is set to the number of cores. If 'auto', n_jobs is set using the function skforecast.utils.select_n_jobs_fit_forecaster. New in version 0.9.0

forecaster_id (str, int)

Name used as an identifier of the forecaster.

Notes

A separate model is created for each forecasting time step. It is important to note that all models share the same parameter and hyperparameter configuration.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
def __init__(
    self, 
    regressor: object,
    steps: int,
    lags: Union[int, np.ndarray, list],
    transformer_y: Optional[object] = None,
    transformer_exog: Optional[object] = None,
    weight_func: Optional[Callable] = None,
    fit_kwargs: Optional[dict] = None,
    n_jobs: Union[int, str] = 'auto',
    forecaster_id: Optional[Union[str, int]] = None,
) -> None:

    self.regressor               = regressor
    self.steps                   = steps
    self.transformer_y           = transformer_y
    self.transformer_exog        = transformer_exog
    self.weight_func             = weight_func
    self.source_code_weight_func = None
    self.last_window             = None
    self.index_type              = None
    self.index_freq              = None
    self.training_range          = None
    self.included_exog           = False
    self.exog_type               = None
    self.exog_dtypes             = None
    self.exog_col_names          = None
    self.X_train_col_names       = None
    self.fitted                  = False
    self.creation_date           = pd.Timestamp.today().strftime('%Y-%m-%d %H:%M:%S')
    self.fit_date                = None
    self.skforecast_version      = skforecast.__version__
    self.python_version          = sys.version.split(" ")[0]
    self.forecaster_id           = forecaster_id

    if not isinstance(steps, int):
        raise TypeError(
            (f"`steps` argument must be an int greater than or equal to 1. "
             f"Got {type(steps)}.")
        )

    if steps < 1:
        raise ValueError(
            f"`steps` argument must be greater than or equal to 1. Got {steps}."
        )

    if not isinstance(n_jobs, int) and n_jobs != 'auto':
        raise TypeError(
            f"`n_jobs` must be an integer or `'auto'`. Got {type(n_jobs)}."
        )

    self.regressors_ = {step: clone(self.regressor) for step in range(1, steps + 1)}
    self.lags = initialize_lags(type(self).__name__, lags)
    self.max_lag = max(self.lags)
    self.window_size = self.max_lag
    self.window_size_diff = self.max_lag

    self.weight_func, self.source_code_weight_func, _ = initialize_weights(
        forecaster_name = type(self).__name__, 
        regressor       = regressor, 
        weight_func     = weight_func, 
        series_weights  = None
    )

    self.fit_kwargs = check_select_fit_kwargs(
                          regressor  = regressor,
                          fit_kwargs = fit_kwargs
                      )

    self.in_sample_residuals = {step: None for step in range(1, steps + 1)}
    self.out_sample_residuals = None

    if n_jobs == 'auto':
        self.n_jobs = select_n_jobs_fit_forecaster(
                          forecaster_name = type(self).__name__,
                          regressor_name  = type(self.regressor).__name__,
                      )
    else:
        self.n_jobs = n_jobs if n_jobs > 0 else cpu_count()

_create_lags(y)

Transforms a 1d array into a 2d array (X) and a 2d array (y). Each row in X is associated with a value of y and it represents the lags that precede it.

Notice that, the returned matrix X_data, contains the lag 1 in the first column, the lag 2 in the second column and so on.

Parameters:

Name Type Description Default
y numpy ndarray

1d numpy ndarray Training time series.

required

Returns:

Name Type Description
X_data numpy ndarray

2d numpy ndarray with the lagged values (predictors). Shape: (samples - max(self.lags), len(self.lags))

y_data numpy ndarray

2d numpy ndarray with the values of the time series related to each row of X_data for each step. Shape: (len(self.steps), samples - max(self.lags))

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
def _create_lags(
    self, 
    y: np.ndarray
) -> Tuple[np.ndarray, np.ndarray]:
    """
    Transforms a 1d array into a 2d array (X) and a 2d array (y). Each row
    in X is associated with a value of y and it represents the lags that
    precede it.

    Notice that, the returned matrix X_data, contains the lag 1 in the first
    column, the lag 2 in the second column and so on.

    Parameters
    ----------
    y : numpy ndarray
        1d numpy ndarray Training time series.

    Returns
    -------
    X_data : numpy ndarray
        2d numpy ndarray with the lagged values (predictors). 
        Shape: (samples - max(self.lags), len(self.lags))
    y_data : numpy ndarray
        2d numpy ndarray with the values of the time series related to each 
        row of `X_data` for each step. 
        Shape: (len(self.steps), samples - max(self.lags))

    """

    n_splits = len(y) - self.max_lag - (self.steps - 1)  # rows of y_data
    if n_splits <= 0:
        raise ValueError(
            (f"The maximum lag ({self.max_lag}) must be less than the length "
             f"of the series minus the number of steps ({len(y) - (self.steps - 1)}).")
        )

    X_data = np.full(shape=(n_splits, len(self.lags)), fill_value=np.nan, dtype=float)
    for i, lag in enumerate(self.lags):
        X_data[:, i] = y[self.max_lag - lag : -(lag + self.steps - 1)] 

    y_data = np.full(shape=(self.steps, n_splits), fill_value=np.nan, dtype=float)
    for step in range(self.steps):
        y_data[step, ] = y[self.max_lag + step : self.max_lag + step + n_splits]

    return X_data, y_data

create_train_X_y(y, exog=None)

Create training matrices from univariate time series and exogenous variables. The resulting matrices contain the target variable and predictors needed to train all the regressors (one per step).

Parameters:

Name Type Description Default
y pandas Series

Training time series.

required
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s. Must have the same number of observations as y and their indexes must be aligned.

`None`

Returns:

Name Type Description
X_train pandas DataFrame

Training values (predictors) for each step. Note that the index corresponds to that of the last step. It is updated for the corresponding step in the filter_train_X_y_for_step method. Shape: (len(y) - self.max_lag, len(self.lags))

y_train dict

Values (target) of the time series related to each row of X_train for each step of the form {step: y_step_[i]}. Shape of each series: (len(y) - self.max_lag, )

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
def create_train_X_y(
    self,
    y: pd.Series,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None
) -> Tuple[pd.DataFrame, dict]:
    """
    Create training matrices from univariate time series and exogenous
    variables. The resulting matrices contain the target variable and predictors
    needed to train all the regressors (one per step).

    Parameters
    ----------
    y : pandas Series
        Training time series.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s. Must have the same
        number of observations as `y` and their indexes must be aligned.

    Returns
    -------
    X_train : pandas DataFrame
        Training values (predictors) for each step. Note that the index 
        corresponds to that of the last step. It is updated for the corresponding 
        step in the filter_train_X_y_for_step method.
        Shape: (len(y) - self.max_lag, len(self.lags))
    y_train : dict
        Values (target) of the time series related to each row of `X_train` 
        for each step of the form {step: y_step_[i]}.
        Shape of each series: (len(y) - self.max_lag, )

    """

    if len(y) < self.max_lag + self.steps:
        raise ValueError(
            (f"Minimum length of `y` for training this forecaster is "
             f"{self.max_lag + self.steps}. Got {len(y)}. Reduce the "
             f"number of predicted steps, {self.steps}, or the maximum "
             f"lag, {self.max_lag}, if no more data is available.")
        )

    check_y(y=y)
    y = input_to_frame(data=y, input_name='y')

    fit_transformer = False if self.fitted else True
    y = transform_dataframe(
            df                = y, 
            transformer       = self.transformer_y,
            fit               = fit_transformer,
            inverse_transform = False,
        )
    y_values, y_index = preprocess_y(y=y)

    if exog is not None:
        check_exog(exog=exog, allow_nan=True)
        exog = input_to_frame(data=exog, input_name='exog')
        if len(exog) != len(y):
            raise ValueError(
                (f"`exog` must have same number of samples as `y`. "
                 f"length `exog`: ({len(exog)}), length `y`: ({len(y)})")
            )
        # Need here for filter_train_X_y_for_step to work without fitting
        self.included_exog = True

        exog = transform_dataframe(
                   df                = exog,
                   transformer       = self.transformer_exog,
                   fit               = fit_transformer,
                   inverse_transform = False
               )

        check_exog_dtypes(exog, call_check_exog=True)

        _, exog_index = preprocess_exog(exog=exog, return_values=False)
        if not (exog_index[:len(y_index)] == y_index).all():
            raise ValueError(
                ("Different index for `y` and `exog`. They must be equal "
                 "to ensure the correct alignment of values.")      
            )

    X_train, y_train = self._create_lags(y=y_values)
    X_train_col_names = [f"lag_{i}" for i in self.lags]
    X_train = pd.DataFrame(
                  data    = X_train,
                  columns = X_train_col_names,
                  index   = y_index[self.max_lag + (self.steps - 1):]
              )

    if exog is not None:
        # Transform exog to match direct format
        # The first `self.max_lag` positions have to be removed from X_exog
        # since they are not in X_lags.
        exog_to_train = exog_to_direct(
                            exog  = exog,
                            steps = self.steps
                        ).iloc[-X_train.shape[0]:, :]
        exog_to_train.index = exog_index[-X_train.shape[0]:]
        X_train = pd.concat((X_train, exog_to_train), axis=1)

    # TODO: move self to fit method and make X_train_col_names a return
    if not self.fitted:
        self.X_train_col_names = X_train.columns.to_list()

    y_train = {step: pd.Series(
                         data  = y_train[step - 1], 
                         index = y_index[self.max_lag + step - 1:][:len(y_train[0])],
                         name  = f"y_step_{step}"
                     )
               for step in range(1, self.steps + 1)}

    return X_train, y_train

filter_train_X_y_for_step(step, X_train, y_train, remove_suffix=False)

Select the columns needed to train a forecaster for a specific step.
The input matrices should be created using create_train_X_y method. This method updates the index of X_train to the corresponding one according to y_train. If remove_suffix=True the suffix "_step_i" will be removed from the column names.

Parameters:

Name Type Description Default
step int

Step for which columns must be selected selected. Starts at 1.

required
X_train pandas DataFrame

Dataframe created with the create_train_X_y method, first return.

required
y_train dict

Dict created with the create_train_X_y method, second return.

required
remove_suffix bool

If True, suffix "_step_i" is removed from the column names.

`False`

Returns:

Name Type Description
X_train_step pandas DataFrame

Training values (predictors) for the selected step.

y_train_step pandas Series

Values (target) of the time series related to each row of X_train. Shape: (len(y) - self.max_lag)

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
def filter_train_X_y_for_step(
    self,
    step: int,
    X_train: pd.DataFrame,
    y_train: dict,
    remove_suffix: bool = False
) -> Tuple[pd.DataFrame, pd.Series]:
    """
    Select the columns needed to train a forecaster for a specific step.  
    The input matrices should be created using `create_train_X_y` method. 
    This method updates the index of `X_train` to the corresponding one 
    according to `y_train`. If `remove_suffix=True` the suffix "_step_i" 
    will be removed from the column names. 

    Parameters
    ----------
    step : int
        Step for which columns must be selected selected. Starts at 1.
    X_train : pandas DataFrame
        Dataframe created with the `create_train_X_y` method, first return.
    y_train : dict
        Dict created with the `create_train_X_y` method, second return.
    remove_suffix : bool, default `False`
        If True, suffix "_step_i" is removed from the column names.

    Returns
    -------
    X_train_step : pandas DataFrame
        Training values (predictors) for the selected step.
    y_train_step : pandas Series
        Values (target) of the time series related to each row of `X_train`.
        Shape: (len(y) - self.max_lag)

    """

    if (step < 1) or (step > self.steps):
        raise ValueError(
            (f"Invalid value `step`. For this forecaster, minimum value is 1 "
             f"and the maximum step is {self.steps}.")
        )

    y_train_step = y_train[step]

    # Matrix X_train starts at index 0.
    if not self.included_exog:
        X_train_step = X_train
    else:
        idx_columns_lags = np.arange(len(self.lags))
        n_exog = (len(self.X_train_col_names) - len(self.lags)) / self.steps
        idx_columns_exog = (
            np.arange((step - 1) * n_exog, (step) * n_exog) + idx_columns_lags[-1] + 1
        )
        idx_columns = np.hstack((idx_columns_lags, idx_columns_exog))
        X_train_step = X_train.iloc[:, idx_columns]

    X_train_step.index = y_train_step.index

    if remove_suffix:
        X_train_step.columns = [col_name.replace(f"_step_{step}", "")
                                for col_name in X_train_step.columns]
        y_train_step.name = y_train_step.name.replace(f"_step_{step}", "")

    return X_train_step, y_train_step

create_sample_weights(X_train)

Crate weights for each observation according to the forecaster's attribute weight_func.

Parameters:

Name Type Description Default
X_train pandas DataFrame

Dataframe created with create_train_X_y and filter_train_X_y_for_step` methods, first return.

required

Returns:

Name Type Description
sample_weight numpy ndarray

Weights to use in fit method.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
def create_sample_weights(
    self,
    X_train: pd.DataFrame,
) -> np.ndarray:
    """
    Crate weights for each observation according to the forecaster's attribute
    `weight_func`.

    Parameters
    ----------
    X_train : pandas DataFrame
        Dataframe created with `create_train_X_y` and filter_train_X_y_for_step`
        methods, first return.

    Returns
    -------
    sample_weight : numpy ndarray
        Weights to use in `fit` method.

    """

    sample_weight = None

    if self.weight_func is not None:
        sample_weight = self.weight_func(X_train.index)

    if sample_weight is not None:
        if np.isnan(sample_weight).any():
            raise ValueError(
                "The resulting `sample_weight` cannot have NaN values."
            )
        if np.any(sample_weight < 0):
            raise ValueError(
                "The resulting `sample_weight` cannot have negative values."
            )
        if np.sum(sample_weight) == 0:
            raise ValueError(
                ("The resulting `sample_weight` cannot be normalized because "
                 "the sum of the weights is zero.")
            )

    return sample_weight

fit(y, exog=None, store_last_window=True, store_in_sample_residuals=True)

Training Forecaster.

Additional arguments to be passed to the fit method of the regressor can be added with the fit_kwargs argument when initializing the forecaster.

Parameters:

Name Type Description Default
y pandas Series

Training time series.

required
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s. Must have the same number of observations as y and their indexes must be aligned so that y[i] is regressed on exog[i].

`None`
store_last_window bool

Whether or not to store the last window of training data.

`True`
store_in_sample_residuals bool

If True, in-sample residuals will be stored in the forecaster object after fitting.

`True`

Returns:

Type Description
None
Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
def fit(
    self,
    y: pd.Series,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None,
    store_last_window: bool = True,
    store_in_sample_residuals: bool = True
) -> None:
    """
    Training Forecaster.

    Additional arguments to be passed to the `fit` method of the regressor 
    can be added with the `fit_kwargs` argument when initializing the forecaster.

    Parameters
    ----------
    y : pandas Series
        Training time series.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s. Must have the same
        number of observations as `y` and their indexes must be aligned so
        that y[i] is regressed on exog[i].
    store_last_window : bool, default `True`
        Whether or not to store the last window of training data.
    store_in_sample_residuals : bool, default `True`
        If `True`, in-sample residuals will be stored in the forecaster object
        after fitting.

    Returns
    -------
    None

    """

    # Reset values in case the forecaster has already been fitted.
    self.index_type          = None
    self.index_freq          = None
    self.last_window         = None
    self.included_exog       = False
    self.exog_type           = None
    self.exog_dtypes         = None
    self.exog_col_names      = None
    self.X_train_col_names   = None
    self.in_sample_residuals = {step: None for step in range(1, self.steps + 1)}
    self.fitted              = False
    self.training_range      = None

    X_train, y_train = self.create_train_X_y(y=y, exog=exog)

    def fit_forecaster(regressor, X_train, y_train, step, store_in_sample_residuals):
        """
        Auxiliary function to fit each of the forecaster's regressors in parallel.

        Parameters
        ----------
        regressor : object
            Regressor to be fitted.
        X_train : pandas DataFrame
            Dataframe created with the `create_train_X_y` method, first return.
        y_train : dict
            Dict created with the `create_train_X_y` method, second return.
        step : int
            Step of the forecaster to be fitted.
        store_in_sample_residuals : bool
            If `True`, in-sample residuals will be stored in the forecaster object
            after fitting.

        Returns
        -------
        Tuple with the step, fitted regressor and in-sample residuals.

        """

        X_train_step, y_train_step = self.filter_train_X_y_for_step(
                                         step          = step,
                                         X_train       = X_train,
                                         y_train       = y_train,
                                         remove_suffix = True
                                     )
        sample_weight = self.create_sample_weights(X_train=X_train_step)
        if sample_weight is not None:
            regressor.fit(
                X             = X_train_step,
                y             = y_train_step,
                sample_weight = sample_weight,
                **self.fit_kwargs
            )
        else:
            regressor.fit(
                X = X_train_step,
                y = y_train_step,
                **self.fit_kwargs
            )

        # This is done to save time during fit in functions such as backtesting()
        if store_in_sample_residuals:
            residuals = (
                (y_train_step - regressor.predict(X_train_step))
            ).to_numpy()

            if len(residuals) > 1000:
                # Only up to 1000 residuals are stored
                rng = np.random.default_rng(seed=123)
                residuals = rng.choice(
                                a       = residuals, 
                                size    = 1000, 
                                replace = False
                            )
        else:
            residuals = None

        return step, regressor, residuals

    results_fit = (
        Parallel(n_jobs=self.n_jobs)
        (delayed(fit_forecaster)
        (
            regressor                 = copy(self.regressor),
            X_train                   = X_train,
            y_train                   = y_train,
            step                      = step,
            store_in_sample_residuals = store_in_sample_residuals
        )
        for step in range(1, self.steps + 1))
    )

    self.regressors_ = {step: regressor 
                        for step, regressor, _ in results_fit}

    if store_in_sample_residuals:
        self.in_sample_residuals = {step: residuals 
                                    for step, _, residuals in results_fit}

    self.fitted = True
    self.fit_date = pd.Timestamp.today().strftime('%Y-%m-%d %H:%M:%S')
    self.training_range = preprocess_y(y=y, return_values=False)[1][[0, -1]]
    self.index_type = type(X_train.index)
    if isinstance(X_train.index, pd.DatetimeIndex):
        self.index_freq = X_train.index.freqstr
    else: 
        self.index_freq = X_train.index.step

    if exog is not None:
        self.included_exog = True
        self.exog_type = type(exog)
        self.exog_dtypes = get_exog_dtypes(exog=exog)
        self.exog_col_names = (
            exog.columns.to_list()
            if isinstance(exog, pd.DataFrame)
            else [exog.name]
        )

    if store_last_window:
        self.last_window = (
            y.iloc[-self.window_size_diff:]
            .copy()
            .to_frame(name=y.name if y.name is not None else 'y')
        )

_create_predict_inputs(steps=None, last_window=None, exog=None)

Create the inputs needed for the prediction process.

Parameters:

Name Type Description Default
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series, pandas DataFrame

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored in self.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`

Returns:

Name Type Description
Xs list

List of numpy arrays with the predictors for each step.

steps list

Steps to predict.

last_window_index pandas Index

Last window Index.

Xs_col_names list

Names of the columns of the matrix created internally for prediction.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
def _create_predict_inputs(
    self,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[Union[pd.Series, pd.DataFrame]] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None
) -> Tuple[list, list, pd.Index, list]:
    """
    Create the inputs needed for the prediction process.

    Parameters
    ----------
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, pandas DataFrame, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in `self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.

    Returns
    -------
    Xs : list
        List of numpy arrays with the predictors for each step.
    steps : list
        Steps to predict.
    last_window_index : pandas Index
        Last window Index.
    Xs_col_names : list
        Names of the columns of the matrix created internally for prediction.

    """

    steps = prepare_steps_direct(
                max_step = self.steps,
                steps    = steps
            )

    if last_window is None:
        last_window = self.last_window

    check_predict_input(
        forecaster_name  = type(self).__name__,
        steps            = steps,
        fitted           = self.fitted,
        included_exog    = self.included_exog,
        index_type       = self.index_type,
        index_freq       = self.index_freq,
        window_size      = self.window_size_diff,
        last_window      = last_window,
        last_window_exog = None,
        exog             = exog,
        exog_type        = self.exog_type,
        exog_col_names   = self.exog_col_names,
        interval         = None,
        alpha            = None,
        max_steps        = None,
        levels           = None,
        series_col_names = None
    )

    last_window = last_window.iloc[-self.window_size_diff:].copy()
    last_window = input_to_frame(data=last_window, input_name='last_window')

    if exog is not None:
        exog = input_to_frame(data=exog, input_name='exog')
        exog = exog.loc[:, self.exog_col_names]
        exog = transform_dataframe(
                   df                = exog,
                   transformer       = self.transformer_exog,
                   fit               = False,
                   inverse_transform = False
               )
        check_exog_dtypes(exog=exog)
        exog_values = exog_to_direct_numpy(
                          exog  = exog.to_numpy()[:max(steps)],
                          steps = max(steps)
                      )[0]
    else:
        exog_values = None

    last_window = transform_dataframe(
                      df                = last_window,
                      transformer       = self.transformer_y,
                      fit               = False,
                      inverse_transform = False
                  )
    last_window_values, last_window_index = preprocess_last_window(
                                                last_window = last_window
                                            )

    X_lags = last_window_values[-self.lags].reshape(1, -1)
    Xs_col_names = [f"lag_{i}" for i in self.lags]
    if exog is None:
        Xs = [X_lags] * len(steps)
    else:
        n_exog = exog.shape[1]
        Xs = [
            np.hstack(
                [X_lags, exog_values[(step - 1) * n_exog : step * n_exog].reshape(1, -1)]
            )
            for step in steps
        ]
        Xs_col_names = Xs_col_names + exog.columns.to_list()

    return Xs, steps, last_window_index, Xs_col_names

create_predict_X(steps=None, last_window=None, exog=None)

Create the predictors needed to predict steps ahead.

Parameters:

Name Type Description Default
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series, pandas DataFrame

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored in self.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`

Returns:

Name Type Description
X_predict pandas DataFrame

Pandas DataFrame with the predictors for each step. The index is the same as the prediction index.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
def create_predict_X(
    self,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[Union[pd.Series, pd.DataFrame]] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None
) -> pd.DataFrame:
    """
    Create the predictors needed to predict `steps` ahead.

    Parameters
    ----------
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, pandas DataFrame, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in `self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.

    Returns
    -------
    X_predict : pandas DataFrame
        Pandas DataFrame with the predictors for each step. The index 
        is the same as the prediction index.

    """

    Xs, steps, last_window_index, Xs_col_names = self._create_predict_inputs(
        steps=steps, last_window=last_window, exog=exog
    )

    idx = expand_index(index=last_window_index, steps=max(steps))
    X_predict = pd.DataFrame(
                    data    = np.concatenate(Xs, axis=0), 
                    columns = Xs_col_names, 
                    index   = idx[np.array(steps) - 1]
                )

    return X_predict

predict(steps=None, last_window=None, exog=None)

Predict n steps ahead.

Parameters:

Name Type Description Default
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series, pandas DataFrame

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored inself.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`

Returns:

Name Type Description
predictions pandas Series

Predicted values.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
def predict(
    self,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[Union[pd.Series, pd.DataFrame]] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None
) -> pd.Series:
    """
    Predict n steps ahead.

    Parameters
    ----------
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, pandas DataFrame, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in` self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.

    Returns
    -------
    predictions : pandas Series
        Predicted values.

    """

    Xs, steps, last_window_index, _ = self._create_predict_inputs(
        steps=steps, last_window=last_window, exog=exog
    )

    regressors = [self.regressors_[step] for step in steps]
    with warnings.catch_warnings():
        # Suppress scikit-learn warning: "X does not have valid feature names,
        # but NoOpTransformer was fitted with feature names".
        warnings.simplefilter("ignore", category=UserWarning)
        predictions = [
            regressor.predict(X)[0] for regressor, X in zip(regressors, Xs)
        ]

    idx = expand_index(index=last_window_index, steps=max(steps))
    predictions = pd.Series(
                      data  = predictions,
                      index = idx[np.array(steps) - 1],
                      name  = 'pred'
                  )

    predictions = transform_series(
                      series            = predictions,
                      transformer       = self.transformer_y,
                      fit               = False,
                      inverse_transform = True
                  )

    return predictions

predict_bootstrapping(steps=None, last_window=None, exog=None, n_boot=500, random_state=123, in_sample_residuals=True, binned_residuals=None)

Generate multiple forecasting predictions using a bootstrapping process. By sampling from a collection of past observed errors (the residuals), each iteration of bootstrapping generates a different set of predictions. See the Notes section for more information.

Parameters:

Name Type Description Default
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored inself.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`
n_boot int

Number of bootstrapping iterations used to estimate predictions.

`500`
random_state int

Sets a seed to the random generator, so that boot predictions are always deterministic.

`123`
in_sample_residuals bool

If True, residuals from the training data are used as proxy of prediction error to create predictions. If False, out of sample residuals are used. In the latter case, the user should have calculated and stored the residuals within the forecaster (see set_out_sample_residuals()).

`True`
binned_residuals Ignored

Not used, present here for API consistency by convention.

None

Returns:

Name Type Description
boot_predictions pandas DataFrame

Predictions generated by bootstrapping. Shape: (steps, n_boot)

Notes

More information about prediction intervals in forecasting: https://otexts.com/fpp3/prediction-intervals.html#prediction-intervals-from-bootstrapped-residuals Forecasting: Principles and Practice (3nd ed) Rob J Hyndman and George Athanasopoulos.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
def predict_bootstrapping(
    self,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[pd.Series] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None,
    n_boot: int = 500,
    random_state: int = 123,
    in_sample_residuals: bool = True,
    binned_residuals: Any = None
) -> pd.DataFrame:
    """
    Generate multiple forecasting predictions using a bootstrapping process. 
    By sampling from a collection of past observed errors (the residuals),
    each iteration of bootstrapping generates a different set of predictions. 
    See the Notes section for more information. 

    Parameters
    ----------
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in` self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.
    n_boot : int, default `500`
        Number of bootstrapping iterations used to estimate predictions.
    random_state : int, default `123`
        Sets a seed to the random generator, so that boot predictions are always 
        deterministic.
    in_sample_residuals : bool, default `True`
        If `True`, residuals from the training data are used as proxy of
        prediction error to create predictions. If `False`, out of sample 
        residuals are used. In the latter case, the user should have
        calculated and stored the residuals within the forecaster (see
        `set_out_sample_residuals()`).
    binned_residuals : Ignored
        Not used, present here for API consistency by convention.

    Returns
    -------
    boot_predictions : pandas DataFrame
        Predictions generated by bootstrapping.
        Shape: (steps, n_boot)

    Notes
    -----
    More information about prediction intervals in forecasting:
    https://otexts.com/fpp3/prediction-intervals.html#prediction-intervals-from-bootstrapped-residuals
    Forecasting: Principles and Practice (3nd ed) Rob J Hyndman and George Athanasopoulos.

    """

    if self.fitted:

        steps = prepare_steps_direct(
                    steps    = steps,
                    max_step = self.steps
                )

        if in_sample_residuals:
            if not set(steps).issubset(set(self.in_sample_residuals.keys())):
                raise ValueError(
                    (f"Not `forecaster.in_sample_residuals` for steps: "
                     f"{set(steps) - set(self.in_sample_residuals.keys())}.")
                )
            residuals = self.in_sample_residuals
        else:
            if self.out_sample_residuals is None:
                raise ValueError(
                    ("`forecaster.out_sample_residuals` is `None`. Use "
                     "`in_sample_residuals=True` or method `set_out_sample_residuals()` "
                     "before `predict_interval()`, `predict_bootstrapping()`, "
                     "`predict_quantiles()` or `predict_dist()`.")
                )
            else:
                if not set(steps).issubset(set(self.out_sample_residuals.keys())):
                    raise ValueError(
                        (f"Not `forecaster.out_sample_residuals` for steps: "
                         f"{set(steps) - set(self.out_sample_residuals.keys())}. "
                         f"Use method `set_out_sample_residuals()`.")
                    )
            residuals = self.out_sample_residuals

        check_residuals = (
            "forecaster.in_sample_residuals" if in_sample_residuals
            else "forecaster.out_sample_residuals"
        )
        for step in steps:
            if residuals[step] is None:
                raise ValueError(
                    (f"forecaster residuals for step {step} are `None`. "
                     f"Check {check_residuals}.")
                )
            elif (residuals[step] == None).any():
                raise ValueError(
                    (f"forecaster residuals for step {step} contains `None` values. "
                     f"Check {check_residuals}.")
                )

    predictions = self.predict(
                      steps       = steps,
                      last_window = last_window,
                      exog        = exog 
                  )

    # Predictions must be in the transformed scale before adding residuals
    predictions = transform_series(
                      series            = predictions,
                      transformer       = self.transformer_y,
                      fit               = False,
                      inverse_transform = False
                  )
    boot_predictions = pd.concat([predictions] * n_boot, axis=1)
    boot_predictions.columns = [f"pred_boot_{i}" for i in range(n_boot)]

    for i, step in enumerate(steps):
        rng = np.random.default_rng(seed=random_state)
        sample_residuals = rng.choice(
                               a       = residuals[step],
                               size    = n_boot,
                               replace = True
                           )
        boot_predictions.iloc[i, :] = boot_predictions.iloc[i, :] + sample_residuals

    if self.transformer_y:
        for col in boot_predictions.columns:
            boot_predictions[col] = transform_series(
                                        series            = boot_predictions[col],
                                        transformer       = self.transformer_y,
                                        fit               = False,
                                        inverse_transform = True
                                    )

    return boot_predictions

predict_interval(steps=None, last_window=None, exog=None, interval=[5, 95], n_boot=500, random_state=123, in_sample_residuals=True, binned_residuals=None)

Bootstrapping based predicted intervals. Both predictions and intervals are returned.

Parameters:

Name Type Description Default
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored inself.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`
interval list

Confidence of the prediction interval estimated. Sequence of percentiles to compute, which must be between 0 and 100 inclusive. For example, interval of 95% should be as interval = [2.5, 97.5].

`[5, 95]`
n_boot int

Number of bootstrapping iterations used to estimate predictions.

`500`
random_state int

Sets a seed to the random generator, so that boot predictions are always deterministic.

`123`
in_sample_residuals bool

If True, residuals from the training data are used as proxy of prediction error to create predictions. If False, out of sample residuals are used. In the latter case, the user should have calculated and stored the residuals within the forecaster (see set_out_sample_residuals()).

`True`
binned_residuals Ignored

Not used, present here for API consistency by convention.

None

Returns:

Name Type Description
predictions pandas DataFrame

Values predicted by the forecaster and their estimated interval.

  • pred: predictions.
  • lower_bound: lower bound of the interval.
  • upper_bound: upper bound of the interval.
Notes

More information about prediction intervals in forecasting: https://otexts.com/fpp2/prediction-intervals.html Forecasting: Principles and Practice (2nd ed) Rob J Hyndman and George Athanasopoulos.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
def predict_interval(
    self,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[pd.Series] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None,
    interval: list = [5, 95],
    n_boot: int = 500,
    random_state: int = 123,
    in_sample_residuals: bool = True,
    binned_residuals: Any = None
) -> pd.DataFrame:
    """
    Bootstrapping based predicted intervals.
    Both predictions and intervals are returned.

    Parameters
    ----------
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in` self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.
    interval : list, default `[5, 95]`
        Confidence of the prediction interval estimated. Sequence of 
        percentiles to compute, which must be between 0 and 100 inclusive. 
        For example, interval of 95% should be as `interval = [2.5, 97.5]`.
    n_boot : int, default `500`
        Number of bootstrapping iterations used to estimate predictions.
    random_state : int, default `123`
        Sets a seed to the random generator, so that boot predictions are always 
        deterministic.
    in_sample_residuals : bool, default `True`
        If `True`, residuals from the training data are used as proxy of
        prediction error to create predictions. If `False`, out of sample 
        residuals are used. In the latter case, the user should have
        calculated and stored the residuals within the forecaster (see
        `set_out_sample_residuals()`).
    binned_residuals : Ignored
        Not used, present here for API consistency by convention.

    Returns
    -------
    predictions : pandas DataFrame
        Values predicted by the forecaster and their estimated interval.

        - pred: predictions.
        - lower_bound: lower bound of the interval.
        - upper_bound: upper bound of the interval.

    Notes
    -----
    More information about prediction intervals in forecasting:
    https://otexts.com/fpp2/prediction-intervals.html
    Forecasting: Principles and Practice (2nd ed) Rob J Hyndman and
    George Athanasopoulos.

    """

    check_interval(interval=interval)

    predictions = self.predict(
                      steps       = steps,
                      last_window = last_window,
                      exog        = exog
                  )

    boot_predictions = self.predict_bootstrapping(
                           steps               = steps,
                           last_window         = last_window,
                           exog                = exog,
                           n_boot              = n_boot,
                           random_state        = random_state,
                           in_sample_residuals = in_sample_residuals
                       )

    interval = np.array(interval) / 100
    predictions_interval = boot_predictions.quantile(q=interval, axis=1).transpose()
    predictions_interval.columns = ['lower_bound', 'upper_bound']
    predictions = pd.concat((predictions, predictions_interval), axis=1)

    return predictions

predict_quantiles(steps=None, last_window=None, exog=None, quantiles=[0.05, 0.5, 0.95], n_boot=500, random_state=123, in_sample_residuals=True, binned_residuals=None)

Bootstrapping based predicted quantiles.

Parameters:

Name Type Description Default
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored inself.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`
quantiles list

Sequence of quantiles to compute, which must be between 0 and 1 inclusive. For example, quantiles of 0.05, 0.5 and 0.95 should be as quantiles = [0.05, 0.5, 0.95].

`[0.05, 0.5, 0.95]`
n_boot int

Number of bootstrapping iterations used to estimate quantiles.

`500`
random_state int

Sets a seed to the random generator, so that boot quantiles are always deterministic.

`123`
in_sample_residuals bool

If True, residuals from the training data are used as proxy of prediction error to create prediction quantiles. If False, out of sample residuals are used. In the latter case, the user should have calculated and stored the residuals within the forecaster (see set_out_sample_residuals()).

`True`
binned_residuals Ignored

Not used, present here for API consistency by convention.

None

Returns:

Name Type Description
predictions pandas DataFrame

Quantiles predicted by the forecaster.

Notes

More information about prediction intervals in forecasting: https://otexts.com/fpp2/prediction-intervals.html Forecasting: Principles and Practice (2nd ed) Rob J Hyndman and George Athanasopoulos.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
def predict_quantiles(
    self,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[pd.Series] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None,
    quantiles: list = [0.05, 0.5, 0.95],
    n_boot: int = 500,
    random_state: int = 123,
    in_sample_residuals: bool = True,
    binned_residuals: Any = None
) -> pd.DataFrame:
    """
    Bootstrapping based predicted quantiles.

    Parameters
    ----------
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in` self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.
    quantiles : list, default `[0.05, 0.5, 0.95]`
        Sequence of quantiles to compute, which must be between 0 and 1 
        inclusive. For example, quantiles of 0.05, 0.5 and 0.95 should be as 
        `quantiles = [0.05, 0.5, 0.95]`.
    n_boot : int, default `500`
        Number of bootstrapping iterations used to estimate quantiles.
    random_state : int, default `123`
        Sets a seed to the random generator, so that boot quantiles are always 
        deterministic.
    in_sample_residuals : bool, default `True`
        If `True`, residuals from the training data are used as proxy of
        prediction error to create prediction quantiles. If `False`, out of
        sample residuals are used. In the latter case, the user should have
        calculated and stored the residuals within the forecaster (see
        `set_out_sample_residuals()`).
    binned_residuals : Ignored
        Not used, present here for API consistency by convention.

    Returns
    -------
    predictions : pandas DataFrame
        Quantiles predicted by the forecaster.

    Notes
    -----
    More information about prediction intervals in forecasting:
    https://otexts.com/fpp2/prediction-intervals.html
    Forecasting: Principles and Practice (2nd ed) Rob J Hyndman and
    George Athanasopoulos.

    """

    check_interval(quantiles=quantiles)

    boot_predictions = self.predict_bootstrapping(
                           steps               = steps,
                           last_window         = last_window,
                           exog                = exog,
                           n_boot              = n_boot,
                           random_state        = random_state,
                           in_sample_residuals = in_sample_residuals
                       )

    predictions = boot_predictions.quantile(q=quantiles, axis=1).transpose()
    predictions.columns = [f'q_{q}' for q in quantiles]

    return predictions

predict_dist(distribution, steps=None, last_window=None, exog=None, n_boot=500, random_state=123, in_sample_residuals=True, binned_residuals=None)

Fit a given probability distribution for each step. After generating multiple forecasting predictions through a bootstrapping process, each step is fitted to the given distribution.

Parameters:

Name Type Description Default
distribution Object

A distribution object from scipy.stats.

required
steps (int, list, None)

Predict n steps. The value of steps must be less than or equal to the value of steps defined when initializing the forecaster. Starts at 1.

  • If int: Only steps within the range of 1 to int are predicted.
  • If list: List of ints. Only the steps contained in the list are predicted.
  • If None: As many steps are predicted as were defined at initialization.
`None`
last_window pandas Series

Series values used to create the predictors (lags) needed to predict steps. If last_window = None, the values stored inself.last_window are used to calculate the initial predictors, and the predictions start right after training data.

`None`
exog pandas Series, pandas DataFrame

Exogenous variable/s included as predictor/s.

`None`
n_boot int

Number of bootstrapping iterations used to estimate predictions.

`500`
random_state int

Sets a seed to the random generator, so that boot predictions are always deterministic.

`123`
in_sample_residuals bool

If True, residuals from the training data are used as proxy of prediction error to create predictions. If False, out of sample residuals are used. In the latter case, the user should have calculated and stored the residuals within the forecaster (see set_out_sample_residuals()).

`True`
binned_residuals Ignored

Not used, present here for API consistency by convention.

None

Returns:

Name Type Description
predictions pandas DataFrame

Distribution parameters estimated for each step.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
def predict_dist(
    self,
    distribution: object,
    steps: Optional[Union[int, list]] = None,
    last_window: Optional[pd.Series] = None,
    exog: Optional[Union[pd.Series, pd.DataFrame]] = None,
    n_boot: int = 500,
    random_state: int = 123,
    in_sample_residuals: bool = True,
    binned_residuals: Any = None
) -> pd.DataFrame:
    """
    Fit a given probability distribution for each step. After generating 
    multiple forecasting predictions through a bootstrapping process, each 
    step is fitted to the given distribution.

    Parameters
    ----------
    distribution : Object
        A distribution object from scipy.stats.
    steps : int, list, None, default `None`
        Predict n steps. The value of `steps` must be less than or equal to the 
        value of steps defined when initializing the forecaster. Starts at 1.

        - If `int`: Only steps within the range of 1 to int are predicted.
        - If `list`: List of ints. Only the steps contained in the list 
        are predicted.
        - If `None`: As many steps are predicted as were defined at 
        initialization.
    last_window : pandas Series, default `None`
        Series values used to create the predictors (lags) needed to 
        predict `steps`.
        If `last_window = None`, the values stored in` self.last_window` are
        used to calculate the initial predictors, and the predictions start
        right after training data.
    exog : pandas Series, pandas DataFrame, default `None`
        Exogenous variable/s included as predictor/s.
    n_boot : int, default `500`
        Number of bootstrapping iterations used to estimate predictions.
    random_state : int, default `123`
        Sets a seed to the random generator, so that boot predictions are always 
        deterministic.
    in_sample_residuals : bool, default `True`
        If `True`, residuals from the training data are used as proxy of
        prediction error to create predictions. If `False`, out of sample 
        residuals are used. In the latter case, the user should have
        calculated and stored the residuals within the forecaster (see
        `set_out_sample_residuals()`).
    binned_residuals : Ignored
        Not used, present here for API consistency by convention.

    Returns
    -------
    predictions : pandas DataFrame
        Distribution parameters estimated for each step.

    """

    boot_samples = self.predict_bootstrapping(
                       steps               = steps,
                       last_window         = last_window,
                       exog                = exog,
                       n_boot              = n_boot,
                       random_state        = random_state,
                       in_sample_residuals = in_sample_residuals
                   )       

    param_names = [p for p in inspect.signature(distribution._pdf).parameters
                   if not p == 'x'] + ["loc", "scale"]
    param_values = np.apply_along_axis(
                       lambda x: distribution.fit(x),
                       axis = 1,
                       arr  = boot_samples
                   )
    predictions = pd.DataFrame(
                      data    = param_values,
                      columns = param_names,
                      index   = boot_samples.index
                  )

    return predictions

set_params(params)

Set new values to the parameters of the scikit learn model stored in the forecaster. It is important to note that all models share the same configuration of parameters and hyperparameters.

Parameters:

Name Type Description Default
params dict

Parameters values.

required

Returns:

Type Description
None
Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
def set_params(
    self, 
    params: dict
) -> None:
    """
    Set new values to the parameters of the scikit learn model stored in the
    forecaster. It is important to note that all models share the same 
    configuration of parameters and hyperparameters.

    Parameters
    ----------
    params : dict
        Parameters values.

    Returns
    -------
    None

    """

    self.regressor = clone(self.regressor)
    self.regressor.set_params(**params)
    self.regressors_ = {step: clone(self.regressor)
                        for step in range(1, self.steps + 1)}

set_fit_kwargs(fit_kwargs)

Set new values for the additional keyword arguments passed to the fit method of the regressor.

Parameters:

Name Type Description Default
fit_kwargs dict

Dict of the form {"argument": new_value}.

required

Returns:

Type Description
None
Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
def set_fit_kwargs(
    self, 
    fit_kwargs: dict
) -> None:
    """
    Set new values for the additional keyword arguments passed to the `fit` 
    method of the regressor.

    Parameters
    ----------
    fit_kwargs : dict
        Dict of the form {"argument": new_value}.

    Returns
    -------
    None

    """

    self.fit_kwargs = check_select_fit_kwargs(self.regressor, fit_kwargs=fit_kwargs)

set_lags(lags)

Set new value to the attribute lags. Attributes max_lag, window_size and window_size_diff are also updated.

Parameters:

Name Type Description Default
lags int, list, numpy ndarray, range

Lags used as predictors. Index starts at 1, so lag 1 is equal to t-1.

  • int: include lags from 1 to lags (included).
  • list, 1d numpy ndarray or range: include only lags present in lags, all elements must be int.
required

Returns:

Type Description
None
Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
def set_lags(
    self, 
    lags: Union[int, list, np.ndarray, range]
) -> None:
    """
    Set new value to the attribute `lags`. Attributes `max_lag`, 
    `window_size` and  `window_size_diff` are also updated.

    Parameters
    ----------
    lags : int, list, numpy ndarray, range
        Lags used as predictors. Index starts at 1, so lag 1 is equal to t-1.

        - `int`: include lags from 1 to `lags` (included).
        - `list`, `1d numpy ndarray` or `range`: include only lags present in 
        `lags`, all elements must be int.

    Returns
    -------
    None

    """

    self.lags = initialize_lags(type(self).__name__, lags)
    self.max_lag = max(self.lags)
    self.window_size = self.max_lag
    self.window_size_diff = self.max_lag

set_out_sample_residuals(residuals, append=True, transform=True, random_state=123)

Set new values to the attribute out_sample_residuals. Out of sample residuals are meant to be calculated using observations that did not participate in the training process.

Parameters:

Name Type Description Default
residuals dict

Dictionary of numpy ndarrays with the residuals of each model in the form {step: residuals}. If len(residuals) > 1000, only a random sample of 1000 values are stored.

required
append bool

If True, new residuals are added to the once already stored in the attribute out_sample_residuals. Once the limit of 1000 values is reached, no more values are appended. If False, out_sample_residuals is overwritten with the new residuals.

`True`
transform bool

If True, new residuals are transformed using self.transformer_y.

`True`
random_state int

Sets a seed to the random sampling for reproducible output.

`123`

Returns:

Type Description
None
Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
def set_out_sample_residuals(
    self, 
    residuals: dict, 
    append: bool = True,
    transform: bool = True,
    random_state: int = 123
) -> None:
    """
    Set new values to the attribute `out_sample_residuals`. Out of sample
    residuals are meant to be calculated using observations that did not
    participate in the training process.

    Parameters
    ----------
    residuals : dict
        Dictionary of numpy ndarrays with the residuals of each model in the
        form {step: residuals}. If len(residuals) > 1000, only a random 
        sample of 1000 values are stored.
    append : bool, default `True`
        If `True`, new residuals are added to the once already stored in the
        attribute `out_sample_residuals`. Once the limit of 1000 values is
        reached, no more values are appended. If False, `out_sample_residuals`
        is overwritten with the new residuals.
    transform : bool, default `True`
        If `True`, new residuals are transformed using self.transformer_y.
    random_state : int, default `123`
        Sets a seed to the random sampling for reproducible output.

    Returns
    -------
    None

    """

    if not isinstance(residuals, dict) or not all(isinstance(x, np.ndarray) for x in residuals.values()):
        raise TypeError(
            (f"`residuals` argument must be a dict of numpy ndarrays in the form "
             "`{step: residuals}`. " 
             f"Got {type(residuals)}.")
        )

    if not self.fitted:
        raise NotFittedError(
            ("This forecaster is not fitted yet. Call `fit` with appropriate "
             "arguments before using `set_out_sample_residuals()`.")
        )

    if self.out_sample_residuals is None:
        self.out_sample_residuals = {step: None 
                                     for step in range(1, self.steps + 1)}

    if not set(self.out_sample_residuals.keys()).issubset(set(residuals.keys())):
        warnings.warn(
            f"""
            Only residuals of models (steps) 
            {set(self.out_sample_residuals.keys()).intersection(set(residuals.keys()))} 
            are updated.
            """
        )

    residuals = {key: value 
                 for key, value in residuals.items() 
                 if key in self.out_sample_residuals.keys()}

    if not transform and self.transformer_y is not None:
        warnings.warn(
            (f"Argument `transform` is set to `False` but forecaster was trained "
             f"using a transformer {self.transformer_y}. Ensure that the new "
             f"residuals are already transformed or set `transform=True`.")
        )

    if transform and self.transformer_y is not None:
        warnings.warn(
            (f"Residuals will be transformed using the same transformer used "
             f"when training the forecaster ({self.transformer_y}). Ensure the "
             f"new residuals are on the same scale as the original time series.")
        )
        for key, value in residuals.items():
            residuals[key] = transform_series(
                                 series            = pd.Series(value, name='residuals'),
                                 transformer       = self.transformer_y,
                                 fit               = False,
                                 inverse_transform = False
                             ).to_numpy()

    for key, value in residuals.items():
        if len(value) > 1000:
            rng = np.random.default_rng(seed=random_state)
            value = rng.choice(a=value, size=1000, replace=False)

        if append and self.out_sample_residuals[key] is not None:
            free_space = max(0, 1000 - len(self.out_sample_residuals[key]))
            if len(value) < free_space:
                value = np.hstack((
                            self.out_sample_residuals[key],
                            value
                        ))
            else:
                value = np.hstack((
                            self.out_sample_residuals[key],
                            value[:free_space]
                        ))

        self.out_sample_residuals[key] = value

get_feature_importances(step, sort_importance=True)

Return feature importance of the model stored in the forecaster for a specific step. Since a separate model is created for each forecast time step, it is necessary to select the model from which retrieve information. Only valid when regressor stores internally the feature importances in the attribute feature_importances_ or coef_. Otherwise, it returns
None.

Parameters:

Name Type Description Default
step int

Model from which retrieve information (a separate model is created for each forecast time step). First step is 1.

required
sort_importance bool

If True, sorts the feature importances in descending order.

True

Returns:

Name Type Description
feature_importances pandas DataFrame

Feature importances associated with each predictor.

Source code in skforecast\ForecasterAutoregDirect\ForecasterAutoregDirect.py
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
def get_feature_importances(
    self, 
    step: int,
    sort_importance: bool = True
) -> pd.DataFrame:
    """
    Return feature importance of the model stored in the forecaster for a
    specific step. Since a separate model is created for each forecast time
    step, it is necessary to select the model from which retrieve information.
    Only valid when regressor stores internally the feature importances in
    the attribute `feature_importances_` or `coef_`. Otherwise, it returns  
    `None`.

    Parameters
    ----------
    step : int
        Model from which retrieve information (a separate model is created 
        for each forecast time step). First step is 1.
    sort_importance: bool, default `True`
        If `True`, sorts the feature importances in descending order.

    Returns
    -------
    feature_importances : pandas DataFrame
        Feature importances associated with each predictor.

    """

    if not isinstance(step, int):
        raise TypeError(
            f"`step` must be an integer. Got {type(step)}."
        )

    if not self.fitted:
        raise NotFittedError(
            ("This forecaster is not fitted yet. Call `fit` with appropriate "
             "arguments before using `get_feature_importances()`.")
        )

    if (step < 1) or (step > self.steps):
        raise ValueError(
            (f"The step must have a value from 1 to the maximum number of steps "
             f"({self.steps}). Got {step}.")
        )

    if isinstance(self.regressor, Pipeline):
        estimator = self.regressors_[step][-1]
    else:
        estimator = self.regressors_[step]

    idx_columns_lags = np.arange(len(self.lags))
    if self.included_exog:
        idx_columns_exog = np.flatnonzero(
                               [name.endswith(f"step_{step}") 
                                for name in self.X_train_col_names]
                           )
    else:
        idx_columns_exog = np.array([], dtype=int)

    idx_columns = np.hstack((idx_columns_lags, idx_columns_exog))
    feature_names = [self.X_train_col_names[i].replace(f"_step_{step}", "") 
                     for i in idx_columns]

    if hasattr(estimator, 'feature_importances_'):
        feature_importances = estimator.feature_importances_
    elif hasattr(estimator, 'coef_'):
        feature_importances = estimator.coef_
    else:
        warnings.warn(
            (f"Impossible to access feature importances for regressor of type "
             f"{type(estimator)}. This method is only valid when the "
             f"regressor stores internally the feature importances in the "
             f"attribute `feature_importances_` or `coef_`.")
        )
        feature_importances = None

    if feature_importances is not None:
        feature_importances = pd.DataFrame({
                                  'feature': feature_names,
                                  'importance': feature_importances
                              })
        if sort_importance:
            feature_importances = feature_importances.sort_values(
                                      by='importance', ascending=False
                                  )

    return feature_importances