pyhctsa.utils.get_dataset¶

pyhctsa.utils.get_dataset(which='e1000')¶

Load predefined datasets for testing and validation.

Parameters:¶

which : str, default="e1000"¶

Dataset identifier.

Options are:

• "e1000": the Empirical1000 dataset of 1000 empirical time series (Fulcher, figshare, doi:10.6084/m9.figshare.5436136). It is not bundled with pyhctsa: the first call downloads it (33 MB) into the user cache directory (~/.cache/pyhctsa; set PYHCTSA_DATA_DIR to use another directory) and verifies its SHA-256 digest.

• "sinusoid": Sinusoidal test data (bundled).

• "noise": Gaussian noise data (T = 1000 sample length time series; bundled).

Returns:¶

List of time series data, where each element is a time series instance as a numpy array.

Return type:¶

list

Raises:¶

ConnectionError – If "e1000" is not cached and cannot be downloaded.