pyamr.metrics.weights

Classes

class pyamr.metrics.weights.SigmoidA(r=200, g=0.5, offset=0.0, scale=1.0, percentiles=None, thresholds=None)[source]

Methods:

`approximated_sigmoid`(x[, x_curves])	This function computes the approximated sigmoid.
`threshold`(r, x[, thresholds])	This function thresholds the r given the values of x.
`weights`(x)	This cuntion computes the weights.
`weights_percentile`(x, percentiles)	This function returns the weights.
`weights_standard`(x)	This function returns the weights.

approximated_sigmoid(x, x_curves=None)[source]

This function computes the approximated sigmoid.

Note: The approximated sigmoid is defined within the interval [0,1].

Parameters:

x (numpy.array) – The values to be converted to weights.
x_curves (tuple) – The values indicating where the low/upper curves should start.

Returns:

r – The weights.

Return type:

numpy.array

threshold(r, x, thresholds=(None, None))[source]

This function thresholds the r given the values of x.

Parameters:

x (numpy.array) – The values to be converted to weights.
r (numpy.array) – The weights to be thresholded.
threshold_low (number;) – The values in x lower will have minimum weight.
treshold_high (number;) – The values in x higher will have maximum weight.