pyhctsa.operations.distribution.outlier_include¶

pyhctsa.operations.distribution.outlier_include(y, threshold_how='abs', inc=0.01)¶

How statistics depend on distributional outliers.

Measures how various statistics of a time series change as more and more outliers are included in the calculation, according to a specified rule for defining outliers.

At each threshold, the mean, standard error, proportion of included points, median, and standard deviation are calculated. Outputs summarize how these statistics change as more extreme points are included.

Parameters:¶

y : array-like¶

The input time series.

threshold_how : {'abs', 'pos', 'neg'}, optional¶

The method for determining outliers:

• ’abs’: Outliers are furthest from the mean (default).

• ’pos’: Outliers are the greatest positive deviations from the mean.

• ’neg’: Outliers are the greatest negative deviations from the mean.

inc : float, optional¶

The increment to move through thresholds (as a fraction of the standard deviation). Default is 0.01.

Returns:¶

Dictionary containing statistics describing how the statistics change as more outliers are included.

Return type:¶

dict