This issue is difficulting my algorithm implementation:

>>> -19.9871/2.8553
-7.0
>>> -19.9871//2.8553
-8.0

The second division should return -7.0 also. I guess this is something related to floating point errors, do you know another way to get the expected result? Using numpy.floor(-19.9871/2.8553) but maybe there is a better solution…