pygsti.extras.drift.signal.renormalizer

Contents

pygsti.extras.drift.signal.renormalizer#

renormalizer(p, method='logistic')#

Takes an arbitrary input vector p and maps it to a vector bounded within [0,1].

  • If method = “sharp”, then it maps any value in p below zero to zero and any value in p above one to one.

  • If method = “logistic” then it ‘squashes’ the vector around it’s mean value using a logistic function. The exact transform for each element x of p is:

    x -> mean - nu + (2*nu) / (1 + exp(-2 * (x - mean) / nu))

    where mean is the mean value of p, and nu is min(1-mean,mean). This transformation is only sensible when the mean of p is within [0,1].

Parameters:

  • p (array of floats) – The vector with values to be mapped to [0,1]

  • method ({'logistic', 'sharp'}) – The method for “squashing” the input vector.

Returns:

The input vector “squashed” using the specified method.

Return type:

numpy.array

Contents