HP Forums

Full Version: XMSTAT returning wrong values for F distribution
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
It looks like FSCP doesn't work so well when both degrees of freedom are odd. I always use the standard test case of F(5;5,5)≈0.05 in order to check algorithms as it's easy to remember and impossible to get the argument order wrong. Also, accuracy of FSDF appears to be poor. Here's the output from an R session:

[attachment=3067]
Reference URL's