gtda.time_series: Time series

The module gtda.time_series implements transformers to preprocess time series or embed them in a higher dimensional space for persistent homology.

Preprocessing

time_series.SlidingWindow([size, stride])

Sliding windows onto the data.

time_series.Resampler([period])

Time series resampling at regular intervals.

time_series.Stationarizer([operation])

Methods for stationarizing time series data.

Time-delay embedding

time_series.TakensEmbedding([time_delay, …])

Point clouds from collections of time series via independent Takens embeddings.

time_series.SingleTakensEmbedding([…])

Representation of a single univariate time series as a point cloud.

time_series.takens_embedding_optimal_parameters(X, …)

Compute the “optimal” parameters for a Takens (time-delay) embedding [1]_ of a univariate time series.

Target preparation

time_series.Labeller([size, func, …])

Target creation from sliding windows over a univariate time series.

Dynamical systems

time_series.PermutationEntropy([n_jobs])

Entropies from sets of permutations arg-sorting rows in arrays.

Multivariate

time_series.PearsonDissimilarity([…])

Pearson dissimilarities from collections of multivariate time series.