You're probably best asking on site, but issues to consider first: (1) clarify what you mean by hand coded, (2) define best in a way that can be operationalzed; (3) failing 2, at least explain what it's supposed to be best at. (4) The question itself is a oncern -- If you can't distinguish uniform from the other two before you collect data you probably should think about the variables more. Are these similar variables to ones anyone has seen before?

(5) You'll need to try a site search or two; somewhat similar questions have been asked before and may already address some of your issues.

Thanks a lot @Glen_b ! I found some related questions, they reminded me of histograms.. I was using normal plots (observations vs natural numbers ), instead of frequency vs observations.

I will be using mean squared error as my metric

hand coded meant, parameter estimation for which we have definitions like, for laplacian distribution, median and mean absolute error being the two parameters.

@ankii Those are details that are worth including as part of a question on the main site.

@mkt umm yeah.. But how else do people estimate say laplacian?

and if it is discussion worthy, I could ask which metric to use. Should I ?