pyhctsa.operations.stationarity.stat_av¶

pyhctsa.operations.stationarity.stat_av(y, what_type='seg', extra_param=5)¶

Simple mean-stationarity metric using the StatAv measure.

This function divides the time series into non-overlapping subsegments, calculates the mean in each segment and returns the standard deviation of this set of means. The method provides a simple way to quantify mean-stationarity in time series data.

For mean-stationary data, the StatAv metric will approach zero, while higher values indicate non-stationarity in the mean.

This implementation is based on [1].

References

Parameters:¶

y : array-like¶

The input time series

what_type : str, optional¶

Method to segment the time series:

• ’seg’: divide into n segments (default)

• ’len’: divide into segments of length n

extra_param : int, optional¶

Specifies either:

• Number of segments when what_type=’seg’ (default=5)

• Segment length when what_type=’len’

Returns:¶

The stat_av statistic. Values closer to zero indicate more stationary means across segments.

Return type:¶

float