postprocessor.core.multisignal.mi.mi

class mi(parameters)

Bases: PostProcessABC

Process to estimate power spectral density (autoregressive model).

Attributes

parameters

Methods

run(signal: pd.DataFrame)

Estimates the mutual information between classes of time series.

Attributes

parameters

Methods

run(signals)

Estimates the mutual information between classes of time series.

as_function
default_parameters

__init__(parameters)

Methods

__init__(parameters)
as_function(data, *extra_data, **kwargs)
default_parameters(*args, **kwargs)
run(signals)	Estimates the mutual information between classes of time series.

Attributes

parameters

run(signals)

Estimates the mutual information between classes of time series.

Uses sklean to optimise a pipeline for classifying the individual time series, choosing the number of PCA components (3-7), the classifier - a support vector machine with either a linear or a radial basis function kernel - and its C and gamma parameters.

Errors are found using bootstrapped datasets.

Parameters: signals: list of pandas.DataFrames

A list of DataFrames. Each DataFrame stores a set of time series, with rows indicating individual time series (e.g. from each cell), and columns indicating time points.

Returns: res: array

Summary statistics from the bootstrapped datasets – the median mutual information and the 10% and 90% confidence limits.

If overtime is True, each row corresponds to a different duration of the time series with the shortest duration, just the first time point, in the first row and the longest duration, the entire time series, in the last row.