I'm trying to do a numeric approach to get (95%) coverage factors for a student distribution depending on the degrees of freedom.
I saw a paper that used an approach
and I wanted to replicate it in here. All seems to work fine when I'm using up to 30000 distributions but above that number something I don't understand happens. The summation in the while loop starts outputting a negative number? I should try to reach 1000000 distributions as the paper says to get reliable coverage factors but there must be an error somewhere that I cannot see.
Thanks for your help!
Solved! Go to Solution.
Well that s because the numeric constant that you are using for shift register of for loop is set to I16.
If you want to iterate to 1000000 distributions set it to U32, it will help you. Do the same for graph also