Releases¶
[0.3.0] - [2021-09-01]¶
Added¶
- New module model_selection_statsmodels to cross-validate, backtesting and grid search AutoReg and SARIMAX models from statsmodels library:
backtesting_autoreg_statsmodels
cv_autoreg_statsmodels
backtesting_sarimax_statsmodels
cv_sarimax_statsmodels
grid_search_sarimax_statsmodels
Changed¶
cv_forecaster
returns cross-validation metrics and cross-validation predictions.- Added an extra column for each parameter in the dataframe returned by
grid_search_forecaster
. - statsmodels 0.12.2 added to requirements
Fixed¶
[0.2.0] - [2021-08-26]¶
Added¶
-
Multiple exogenous variables can be passed as pandas DataFrame.
-
Documentation at https://joaquinamatrodrigo.github.io/skforecast/
-
New unit test
-
Increased typing
Changed¶
- New implementation of
ForecasterAutoregMultiOutput
. The training process in the new version creates a different X_train for each step. See Direct multi-step forecasting for more details. Old versión can be acces withskforecast.deprecated.ForecasterAutoregMultiOutput
.
Fixed¶
[0.1.9] - 2121-07-27¶
Added¶
-
Logging total number of models to fit in
grid_search_forecaster
. -
Class
ForecasterAutoregCustom
. -
Method
create_train_X_y
to facilitate access to the training data matrix created fromy
andexog
.
Changed¶
-
New implementation of
ForecasterAutoregMultiOutput
. The training process in the new version creates a different X_train for each step. See Direct multi-step forecasting for more details. Old versión can be acces withskforecast.deprecated.ForecasterAutoregMultiOutput
. -
Class
ForecasterCustom
has been renamed toForecasterAutoregCustom
. However,ForecasterCustom
will still remain to keep backward compatibility. -
Argument
metric
incv_forecaster
,backtesting_forecaster
,grid_search_forecaster
andbacktesting_forecaster_intervals
changed from 'neg_mean_squared_error', 'neg_mean_absolute_error', 'neg_mean_absolute_percentage_error' to 'mean_squared_error', 'mean_absolute_error', 'mean_absolute_percentage_error'. -
Check if argument
metric
incv_forecaster
,backtesting_forecaster
,grid_search_forecaster
andbacktesting_forecaster_intervals
is one of 'mean_squared_error', 'mean_absolute_error', 'mean_absolute_percentage_error'. -
time_series_spliter
doesn't include the remaining observations in the last complete fold but in a new one whenallow_incomplete_fold=True
. Take in consideration that incomplete folds with few observations could overestimate or underestimate the validation metric.
Fixed¶
- Update lags of
ForecasterAutoregMultiOutput
aftergrid_search_forecaster
.
[0.1.8.1] - 2021-05-17¶
Added¶
set_out_sample_residuals
method to store or update out of sample residuals used bypredict_interval
.
Changed¶
-
backtesting_forecaster_intervals
andbacktesting_forecaster
print number of steps per fold. -
Only stored up to 1000 residuals.
-
Improved verbose in
backtesting_forecaster_intervals
.
Fixed¶
-
Warning of inclompleted folds when using
backtesting_forecast
with aForecasterAutoregMultiOutput
. -
ForecasterAutoregMultiOutput.predict
allow exog data longer than needed (steps). -
backtesting_forecast
prints correctly the number of folds when remainder observations are cero. -
Removed named argument X in
self.regressor.predict(X)
to allow using XGBoost regressor. -
Values stored in
self.last_window
when trainingForecasterAutoregMultiOutput
.
[0.1.8] - 2021-04-02¶
Added¶
- Class
ForecasterAutoregMultiOutput.py
: forecaster with direct multi-step predictions. - Method
ForecasterCustom.predict_interval
andForecasterAutoreg.predict_interval
: estimate prediction interval using bootstrapping. skforecast.model_selection.backtesting_forecaster_intervals
perform backtesting and return prediction intervals.
Changed¶
Fixed¶
[0.1.7] - 2021-03-19¶
Added¶
- Class
ForecasterCustom
: same functionalities asForecasterAutoreg
but allows custom definition of predictors.
Changed¶
grid_search forecaster
adapted to work with objectsForecasterCustom
in addition toForecasterAutoreg
.
Fixed¶
[0.1.6] - 2021-03-14¶
Added¶
- Method
get_feature_importances
toskforecast.ForecasterAutoreg
. - Added backtesting strategy in
grid_search_forecaster
. - Added
backtesting_forecast
toskforecast.model_selection
.
Changed¶
- Method
create_lags
return a matrix where the order of columns match the ascending order of lags. For example, column 0 contains the values of the minimum lag used as predictor. - Renamed argument
X
tolast_window
in methodpredict
. - Renamed
ts_cv_forecaster
tocv_forecaster
.
Fixed¶
[0.1.4] - 2021-02-15¶
Added¶
- Method
get_coef
toskforecast.ForecasterAutoreg
.