If X ~ F_{(m,n) }and I'd like to get Y~Beta_{(m/2,n/2). }Should I let Y= (mX/n) / (1 + (mX/n))? If so, what are the next steps?

Printable View

- Jul 14th 2012, 08:05 AMcwertyHow to derive beta distribution from F distribution?
If X ~ F

_{(m,n) }and I'd like to get Y~Beta_{(m/2,n/2). }Should I let Y= (mX/n) / (1 + (mX/n))? If so, what are the next steps?