How can I avoid the dreaded P=NaN when using chi2gof?

I can test a column vector using chi2gof to determine the fit to a specific distribution and Matlab will return a P value of NaN. I can test the exact same data in Minitab and get a real result. I know the issue has to do with how Matlab determines the degrees of freedom and that number being too low. What I don't understand is how to work around this problem. Minitab clearly does.

Answer 1

[Nikhil Vyas]

This answer might help you:

http://www.mathworks.com/matlabcentral/answers/55082-chi-square-test-with-2-bins-why-do-i-get-p-nan#answer_66878