# How to compare goodness of fit between gamma mixture and beta mixture in R?

I have data presenting two peaks that I fitted with both a mixture of gamma and a mixture of beta distributions, in R. The data have for support [0,1]. How to assess which mixture (gamma vs beta) is best?

• This is strange, gamma distribution is supported on $(0,\infty)$ while beta is supported on $(0,1)$. Or do you mean something else? Please clarify! – kjetil b halvorsen May 6 '17 at 21:41
• Thank you for your comment. I edited the post to indicate that the data have for support [0,1]. So even though the beta distribution is more naturally suited to the problem, I am still interested in comparing the fits obtained from a beta mixture and a gamma mixture. – michael May 8 '17 at 9:35