Matrices "belongs" function is not giving the expected result on non-square matrices

[elodiemaignant]

I think it has to do with the use of "&" logic operator rather than "and" one.

import geomstats.backend as gs
import geomstats.datasets.utils as data_utils
from geomstats.geometry.matrices import Matrices

M = Matrices(3,3)

N = gs.zeros((3,3))
print(M.belongs(N))

M = Matrices(3,2)

N = gs.zeros((3,2))
print(M.belongs(N))

[ninamiolane]

Hi @emaignant , thank you for submitting this issue! As a participant to the challenge, feel free to mention the issues you have submitted as part of your notebook within the section "Describe the limitations of the packages Geomstats and Giotto-tda".

Which python version and geomstats version do you use? It was working in my environment for python 3.6, 3.7, 3.8 and 3.9, I was not able to reproduce the bug.

I did replace the & by and in the codebase, as you suggested, and added unit tests on geomstats - see this PR: https://github.com/geomstats/geomstats/pull/942

Let me know if this solves the issue on your side? Note that you will have to pull geomstats' master branch from GitHub, geomstats PyPi version will be updated in a few days only.

[nguigs]

For the record: https://stackoverflow.com/questions/22646463/and-boolean-vs-bitwise-why-difference-in-behavior-with-lists-vs-nump

[ninamiolane]

@emaignant was your issue solved?

[ninamiolane]

Solved, as per slackcommunications with @emaignant