F distribution on the 41C

[Dieter]

(12-19-2015 02:37 PM)Namir Wrote:  Thanks Dieter for the observation. I found the bug:

If Df2 > 1 And Df1 > 0 Then

And changed it (in the original thread) to:

If Df2 > 1 And Df1 > 1 Then

No I get the same result for Fcfd(1,5,3).

Yes, this example now returns the correct result.
But then try df2=1 and df1 odd. The problem persists:
The CDF for x=3, df1=5, df2=1 is returned as 0,839 while the true result is 0,588...

Instead of all these nested if-then-else-elseif constructions for odd and even dof and various complicated trig-power-series: why don't you simply use the regularized Beta function? Simple, reliable, straightforward and easy to implement.

Dieter