API¶
hcrystalball.model_selection Package¶
Functions¶
|
Extends gridsearch with provided model. |
|
Filter provided dataframe by {column:value} rules. |
|
Get grid search object based on selection criteria. |
|
Load information about stored model selection |
|
Partition data by values found in one or more columns. |
|
Partition data by one column and a fixed set ov values within that column. |
|
Prepare data for model selection. |
|
Run parallel cross validation on data and select best model |
|
Find the best model from grid_search for each partition |
|
Run cross validation on data and select best model |
Classes¶
|
Time series cross-validator. |
|
Enable large scale cross validation easily accessible. |
|
Consolidate infromation/methods from cross validation for 1 time series |
Variables¶
|
Returns unpickled file or json metadata from directory. |
|
Returns unpickled file or json metadata from directory. |
hcrystalball.wrappers Package¶
Functions¶
|
Factory function returning the model specific SklearnWrapper with provided |
Classes¶
|
Wrapper for fbprophet.Prophet model |
|
Wrapper for |
|
Wrapper for |
|
Wrapper for |
|
|
|
Wrapper for TBATS model |
|
Wrapper for BATS model |
hcrystalball.ensemble Package¶
Classes¶
|
SimpleEnsemble model, which takes a list of any hcrystalball model wrapper instance(s) as base learners and aggregates their prediction using |
|
StackingEnsemble model, which takes a list of any hcrystalball model wrapper instance(s) as base learners. |
hcrystalball.feature_extraction Package¶
Classes¶
|
Generate holiday feature based on provided ISO code |
|
Generate seasonal feature columns using one-hot encoding. |
hcrystalball.compose Package¶
Classes¶
|
Time Series compatible ColumnTransformer. |
hcrystalball.preprocessing Package¶
Classes¶
|
Enable transformation of the target. |
hcrystalball.metrics Package¶
Functions¶
|
Get a scorer supporting storing data for gridsearch from string. |
|
Make a scorer from a performance metric or loss function. |
hcrystalball.utils Module¶
Functions¶
|
Get the string representation of a model |
|
Load subset of Rossmann store sales dataset. |
|
Generate a unique hash for a model using its string representation |
|
Provide easy way how to generate dummy data for tests or tutorial purposes |
|
Generate a unique hash for data partition |
|
Generate dummy daily time series data compatible with hcrystalball API. |
|
Imports optional dependencies. |