Closed BryceStevenWilley closed 2 years ago
Realized that we're not able to generate the exact same sequence of samples, just because we're calling the random function in different orders, but on the civil docketing statement, I'm getting a mean of 219-221 and a standard dev of 91.8-92.9 both before and after this change, no discernible difference in output. So the final results should be unchanged by this PR.
Makes the monte carlo a bit faster by using numpy to generate an N sized array instead of one at a time. It probably starts being noticably faster than the original code at ~5000 samples.