Time series transformations

The sktime.transformations module contains classes for data transformations.

Panel transformers

Dictionary-based

PAA([num_intervals])	(PAA) Piecewise Aggregate Approximation Transformer, as described in Eamonn Keogh, Kaushik Chakrabarti, Michael Pazzani, and Sharad Mehrotra.
SFA([word_length, alphabet_size, …])	Symbolic Fourier Approximation (SFA) Transformer.
SAX([word_length, alphabet_size, …])	SAX (Symbolic Aggregate approXimation) transformer.

Summarize

DerivativeSlopeTransformer()	Derivative slope transformer.
PlateauFinder([value, min_length])	Plateau finder transformer.
RandomIntervalFeatureExtractor([…])	Random interval feature extractor transform.
FittedParamExtractor(forecaster, param_names)	Fitted parameter extractor.

tsfresh

TSFreshRelevantFeatureExtractor([…])	Transformer for extracting and selecting features.
TSFreshFeatureExtractor([…])	Transformer for extracting time series features.

Catch22

Catch22([outlier_norm, replace_nans, n_jobs])

Canonical Time-series Characteristics (catch22).

Compose

ColumnTransformer(transformers[, remainder, …])	Column-wise application of transformers.
ColumnConcatenator()	Concatenate multivariate series to a long univariate series.
SeriesToSeriesRowTransformer(transformer[, …])	Series-to-series row transformer.
SeriesToPrimitivesRowTransformer(transformer)	Series-to-primitives row transformer.

make_row_transformer(transformer[, …])

Cate InstanceTransformer based on transform type, factory function.

Matrix profile

MatrixProfile([m])

Return the matrix profile and index profile for each time series of a dataset.

PCA

PCATransformer([n_components])

Principal Components Analysis applied to panel of time seires.

Reduce

Tabularizer()

A transformer that turns time series/panel data into tabular data.

Rocket

Rocket([num_kernels, normalise, n_jobs, …])	ROCKET.
MiniRocket([num_kernels, …])	MINIROCKET.
MiniRocketMultivariate([num_kernels, …])	MINIROCKET (Multivariate).

Segment

IntervalSegmenter([intervals])	Interval segmentation transformer.
RandomIntervalSegmenter([n_intervals, …])	Random interval segmenter transformer.

Signature

SignatureTransformer([augmentation_list, …])

Transformation class from the signature method.

Series transformers

Detrend

Detrender([forecaster])	Remove a trend from a series.
Deseasonalizer([sp, model])	Remove seasonal components from a time series.
ConditionalDeseasonalizer([…])	Remove seasonal components from time series, conditional on seasonality test.
STLTransformer([sp, seasonal, trend, …])	Remove seasonal components from a time-series using STL.

Adapt

TabularToSeriesAdaptor(transformer[, …])

Adapt scikit-learn-like transformations to time series setting.

Box-Cox

BoxCoxTransformer([bounds, method, sp])	Box-Cox power transform.
LogTransformer()	Natural logarithm transformation.

Scaled Logit

ScaledLogitTransformer([lower_bound, …])

Scaled logit transform or Log transform.

ClaSP

ClaSPTransformer([window_length, scoring_metric])

ClaSP (Classification Score Profile) Transformer.

Difference

Differencer([lags, drop_na, na_handling])

Apply iterative differences to a timeseries.

Auto-correlation

AutoCorrelationTransformer([adjusted, …])	Auto-correlation transformer.
PartialAutoCorrelationTransformer([n_lags, …])	Partial auto-correlation transformer.

Cosine

CosineTransformer()

Cosine transformation.

Exponent

ExponentTransformer([power, offset])	Apply element-wise exponentiation transformation to a time series.
SqrtTransformer([offset])	Apply element-sise square root transformation to a time series.

Matrix Profile

MatrixProfileTransformer([window_length])

Calculate the matrix profile of a time series.

Missing value imputation

Imputer([method, random_state, value, …])

Missing value imputation.

Datetime feature generation

DateTimeFeatures([ts_freq, feature_scope, …])

DateTime Feature Extraction for use in e.g.

Lagged Window Summarizer

WindowSummarizer([lag_config, lag_feature, …])

Transformer for extracting time series features.

Outlier detection

HampelFilter([window_length, n_sigma, k, …])

Use HampelFilter to detect outliers based on a sliding window.

Composition

TransformerPipeline(steps)	Pipeline of transformers compositor.
FeatureUnion(transformer_list[, n_jobs, …])	Concatenates results of multiple transformer objects.
FitInTransform(transformer[, …])	Transformer composition to always fit a given transformer on the transform data only.

OptionalPassthrough(transformer[, passthrough])	Wrap an existing transformer to tune whether to include it in a pipeline.
ColumnwiseTransformer(transformer[, columns])	Apply a transformer columnwise to multivariate series.

Theta

ThetaLinesTransformer([theta])

Decompose the original data into two or more Theta-lines.

Augmenter

InvertAugmenter()	Augmenter inverting the time series by multiplying it by -1.
RandomSamplesAugmenter([n, …])	Draw random samples from time series.
ReverseAugmenter()	Augmenter reversing the time series.
WhiteNoiseAugmenter([scale, random_state])	Augmenter adding Gaussian (i.e.

Summary

SummaryTransformer([summary_function, quantiles])

Calculate summary value of a time series.

FeatureSelection

FeatureSelection([method, n_columns, …])

Select exogenous features.

STLBootstrapTransformer

STLBootstrapTransformer(n_series, sp, …)

Creates a population of similar time series.

MovingBlockBootstrapTransformer

MovingBlockBootstrapTransformer(n_series, …)

Moving Block Bootstrapping method for synthetic time series generation.