CODEWITHSUNDEEP
pythonnumpymatplotlibindexingscipy
SFlowtY
SFlowtY

Reputation: 47

cross_from_above deprecated in future matplotlib... replacement function?

It seems that matplotlib will deprecate the cross_from_above and cross_from_below functions in the upcoming version 3.1

This is a shame as they're very useful tools for "returning the indices where a 1D array crosses a threshold from above/below". See current documentation at: https://matplotlib.org/api/mlab_api.html

I can't find any discussion of this online so wonder if there are replacement functions that I should be using instead for the same functionality?

Upvotes: 2

Views: 56

Answers (1)

ImportanceOfBeingErnest
ImportanceOfBeingErnest

Reputation: 339580

There is no replacement for these functions. But they aren't very complicated. Here is the literal copy of the source code. 

def cross_from_above(x, threshold):
    """
    return the indices into *x* where *x* crosses some threshold from above.
    """
    x = np.asarray(x)
    ind = np.nonzero((x[:-1] >= threshold) & (x[1:] < threshold))[0]
    if len(ind):
        return ind+1
    else:
        return ind

and 

def cross_from_below(x, threshold):
    """
    return the indices into *x* where *x* crosses some threshold from below.
    """
    x = np.asarray(x)
    ind = np.nonzero((x[:-1] < threshold) & (x[1:] >= threshold))[0]
    if len(ind):
        return ind+1
    else:
        return ind

where np is numpy.

Essentially they both contain a single line of code, which is easy to replicate or modified in any actual use case.

Upvotes: 2

Related Questions