ConstantDetection.md

Constant Detection

A collection of quality check routines to find constant values and plateaus.

Index

• constant
• constants_varianceBased

constant

constant(window, thresh=0)

parameter	data type	default value	description
window	integer/offset string		Minimum count or a duration values need to identical to become plateau candidates. See condition (1)
thresh	float	0	Maximum difference between values to still consider them constant. See condition (2)

This functions flags plateaus/series of constant values of length window if their difference is smaller than thresh.

A set of consecutive values x_n, ..., x_{n+k} of a time series x_t is considered to be constant, if:

1. k \ge window
2. |x_n - x_{n+s}| \le thresh, s \in {1,2, ..., k}

constants_varianceBased

constants_varianceBased(window="12h", thresh=0.0005,
                        max_missing=Inf, max_consec_missing=Inf)

parameter	data type	default value	description
window	offset string		Minimum duration during which values need to identical to become plateau candidates. See condition (1)
thresh	float	0.0005	Maximum variance of a group of values to still consider them constant. See condition (2)
max_missing	integer	None	Maximum number of missing values allowed in window, by default this condition is ignored
max_consec_missing	integer	None	Maximum number of consecutive missing values allowed in window, by default this condition is ignored

This function flags plateaus/series of constant values. Any set of consecutive values x_n,..., x_{n+k} of a timeseries x_t is flagged, if:

1. k \ge window
2. \sigma(x_n,..., x_{n+k}) \le thresh

NOTE:

• Only works for time series
• The time series is expected to be harmonized to an equidistant frequency grid
• When max_missing or max_consec_missing are set, plateaus not fulfilling the respective condition will not be flagged