pyValEIA.utils.filters

Functions to filter data.

Functions

simple_barrel_roll(xvar, yvar, barrel_radius[, ...])	Roll barrel over data to detrended over large decreases in value.
rolling_nanmeasure(arr, window[, measure])	Calculate the rolling mean or median of an array with or without nans.
find_nan_ranges(arr)	Identify continuous ranges of NaN values in an array.
find_all_gaps(inds)	Identify gaps in a list of indices.
detect_outliers(arr)	Detect outliers in an array.

Module Contents

pyValEIA.utils.filters.simple_barrel_roll(xvar, yvar, barrel_radius, envelope=True, envelope_lower=0.6, envelope_upper=0.2)

Roll barrel over data to detrended over large decreases in value.

Parameters:

xvar: array-like

Independent data variable along which yvar will be smoothed.

yvar: array-like

Dependent data variable of the same size as xvar; needs to be scaled such that its magnitude is similar to yvar

barrel_radiusdouble

Radius of the ‘barrel’ rolling over the data in the same units as xvar

envelopebool

If True, the barrel roll results will be used as constraints for the original yvar values (default=True)

envelope_lowerfloat

Fraction starting from zero to multiply the minimum by, used to calculate the lower limit of the barrel roll envelope; a larger value creates a larger lower envelope (default=0.6)

envelope_upperfloat

Fraction starting from zero to multiply the maximum by, used to calculate the upper limit of the barrel roll envelope; a larger value creates a larger upper envelope (default=0.2)

Returns:

yvar_detarray-like

Detrended values for the dependent variable, yvar

pyValEIA.utils.filters.rolling_nanmeasure(arr, window, measure='mean')

Calculate the rolling mean or median of an array with or without nans.

Parameters:

arr: array-like

array of values to roll over

windowint

window size

measurestr

Method to apply to data; ‘mean’, ‘median’, ‘average’ (default=’mean’)

Returns:

outarray-like

rolling measured array of same length as original

pyValEIA.utils.filters.find_nan_ranges(arr)

Identify continuous ranges of NaN values in an array.

Parameters:

arrarray-like

array with nans

Returns:

nan_listlist-like

List of (start_idx, end_idx) for each contiguous NaN section

pyValEIA.utils.filters.find_all_gaps(inds)

Identify gaps in a list of indices.

Parameters:

indslist-like

Array of indices

Returns:

gap_indiceslist-like

Indices of gap start and end

Notes

For example, in an array of arr=[2,3,5,6,7,8], this function will return gap_inds=[1] to indicate where the gap starts.

pyValEIA.utils.filters.detect_outliers(arr)

Detect outliers in an array.

Parameters:

arrarray-like

Array of numbers

Returns:

outlier_indicesarray-like

array of indices where arr has outliers

Notes

Uses InterQuartile Range (IQR) IQR = q3 - q1 outlier > q3 + 1.5 * IQR outlier < q1 - 1.5 * IQR