pyhctsa.operations.nonlinearity.embed_pca¶

pyhctsa.operations.nonlinearity.embed_pca(y, tau='ac', m=3)¶

Reconstructs the time series as a time-delay embedding, and performs Principal Components Analysis on the result. This technique is known as singular spectrum analysis [1].

References

Parameters:¶

y : array-like¶

Input time series.

tau : str or int¶

The time-delay, can be an integer or ‘ac’, or ‘mi’ for first zero-crossing of the autocorrelation function or first minimum of the automutual information, respectively. Default is 'ac'.

m : int¶

The embedding dimension. Default is 3.

Returns:¶

Various statistics summarizing the obtained eigenvalue distribution.

Return type:¶

dict