Writing your own 'numpy' functions is a real easy way to get confused

[malramsay64]

Because numpy is so ubiquitous, particularly within the scientific space, writing your own functions which replicate some of their functionality is a great way to confuse maintainers or anyone looking through your code.

Something like stats or statistics would be a much clearer choice. I will mention there is a statistics module in the python3 standard library.

[mpurg]

Good point, I've switched to stats.
I make use of the statistics functions in Py3.
(branch devel)