[ ] create a test to test the squarepoissond function provided by Dr Wright to see if it produces the same result as the matlab version
[ ] the test should pass to be considered complete meaning the results should be fairly close (not exact due to differences in accuracy between the 2 languages
created a testing class to test the results of the squarepoisson function given by Dr Wright
got data from matlab and save it to files to then load into the python code for comparision
having issues with the fh variable being different between the 2. looking into the matlab code and comparing it to the python to see if the code pieces are doing the same for fh
it looks like there are some operation happening in matlab that aren't happening in python. Started working on adding these operation to see if that fixes the problem