static_covariates_transformer

Generic wrapper class for scalers/encoders/transformers of static covariates. This transformer acts only on static covariates of the series passed to fit(), transform(), fit_transform(), and inverse_transform(). It can both scale numerical features, as well as encode categorical features.

The underlying transformer_num and transformer_cat have to implement the fit(), transform(), and inverse_transform() methods (typically from scikit-learn).

By default, numerical and categorical columns/features are inferred and allocated to transformer_num and transformer_cat, respectively. Alternatively, specify which columns to scale/transform with cols_num and cols_cat.

Both transformer_num and transformer_cat are fit globally on static covariate data from all series passed to anomx.data.processing.transformers.StaticCovariatesTransformer.fit

Parameters

transformer_num

The transformer to transform numeric static covariate columns with. It must provide fit(), transform() and inverse_transform() methods. Default: sklearn.preprocessing.MinMaxScaler with feature_range=(0, 1); this will scale all values between 0 and 1.

transformer_cat

The encoder to transform categorical static covariate columns with. It must provide fit(), transform() and inverse_transform() methods. Default: sklearn.preprocessing.OrdinalEncoder; this will convert categories into integer valued arrays where each integer stands for a specific category.

cols_num

Optionally, a list of column names for which to apply the numeric transformer transformer_num. By default, the transformer will infer all numerical features based on types, and scale them with transformer_num. If an empty list, no column will be scaled.

cols_cat

Optionally, a list of column names for which to apply the categorical transformer transformer_cat. By default, the transformer will infer all categorical features based on types, and transform them with transformer_cat. If an empty list, no column will be transformed.

name

A specific name for the anomx.data.processing.transformers.StaticCovariatesTransformer.

n_jobs

The number of jobs to run in parallel. Parallel jobs are created only when a Sequence[TimeSeries] is passed as input to a method, parallelising operations regarding different TimeSeries. Defaults to 1 (sequential). Setting the parameter to -1 means using all the available processors. Note: for a small amount of data, the parallelisation overhead could end up increasing the total required amount of time.

verbose

Optionally, whether to print operations progress

Examples

>>> import numpy as np
>>> import pandas as pd
>>> from anomx.data.timeseries import TimeSeries
>>> from anomx.data.processing.transformers import StaticCovariatesTransformer
>>> static_covs = pd.DataFrame(data={"num": [0, 2, 1], "cat": ["a", "c", "b"]})
>>> series = TimeSeries.from_values(
>>>     values=np.random.random((10, 3)),
>>>     columns=["comp1", "comp2", "comp3"],
>>>     static_covariates=static_covs,
>>> )
>>> transformer = StaticCovariatesTransformer()
>>> series_transformed = transformer.fit_transform(series)
>>> print(series.static_covariates)
static_covariates  num cat
component
comp1               0.0   a
comp2               2.0   c
comp3               1.0   b
>>> print(series_transformed.static_covariates)
static_covariates  num  cat
component
comp1               0.0  0.0
comp2               1.0  2.0
comp3               0.5  1.0