I brute force it up to n sample points. For the first image n was 100. Yes, I allow the result to be near a gaussian integer. For the first image I counted eve

I brute force it up to n sample points. For the first image n was 100. Yes, I allow the result to be near a gaussian integer. For the first image I counted everything as a gaussian integer where (real%1<0.1 && imaginary%1<0.1)