pretab.transformers.RollingStatsTransformer

class pretab.transformers.RollingStatsTransformer(window_size=5, stats=('mean', 'std'))

Bases: BasePreTabTransformer

Compute rolling-window statistics over time-series inputs.

A sliding window of fixed size is moved across each feature and the requested summary statistics are computed within each window.

Parameters:

• window_size (int, default=5) – Number of consecutive observations in each rolling window.

• stats (tuple of str, default=("mean", "std")) – Statistics to compute. Any of "mean", "std", "min", "max".

Notes

Using a sliding window of size window_size yields n_samples - window_size + 1 output rows. Each requested statistic adds one column per input feature.

This is a standalone time-series utility. It intentionally changes the row count and assumes the rows are ordered in time, so it does not satisfy the row-count-preserving contract that ColumnTransformer (and therefore Preprocessor) require. Apply it directly to an ordered array rather than routing it through the preprocessing pipeline.

Examples

>>> import numpy as np
>>> from pretab.transformers import RollingStatsTransformer
>>> X = np.arange(10).reshape(-1, 1).astype(float)
>>> transformer = RollingStatsTransformer(window_size=3, stats=("mean", "std"))
>>> transformer.fit_transform(X).shape
(8, 2)

__init__(window_size=5, stats=('mean', 'std'))

Methods

__init__([window_size, stats])
fit(X[, y])
fit_transform(X[, y])	Fit to data, then transform it.
get_feature_names_out([input_features])	Return output feature names of the form {feature}_{suffix}{j}.
get_metadata_routing()	Get metadata routing of this object.
get_params([deep])	Get parameters for this estimator.
set_output(*[, transform])	Set output container.
set_params(**params)	Set the parameters of this estimator.
transform(X)

Attributes

total_output_dim_	Total number of output columns produced across all input features.
n_features_in_
adaptive