I have a question, for my bachelor's thesis I wrote a program that classifies what kind of distribution a QQ resembles most, I trained the model using synthetic data: I created the QQ plots by using the built in functions in R e.g. qqnorm(rnorm(100)) etc. I would like to test the model using real world data in order to do this I have to know what the hypothesized distribution of the data is and label the QQ plots accordingly. I am looking for datasets that may have the following distributions: normal, student t, gamma, chisq, uniform, poisson, binomial.
If you think that there's a different way to evaluate my model please let me now!