x = qf(p,df1,df2)
Input p probability
df1,df2 degrees of freedom
(p,df1,df2 can be scalar or matrix with common size)
Output F inverse at p of the F cumulative distribution
function with degrees of freedom df1, df2
(the F law with degrees of freedom df1 and df2
is the law of (X/df1)/(Y/df2) where X and Y
are independent Chisquare random variables
with one degree of freedom)