Dear Clyde,
could you perhaps expand on your idea to use the - cmp - command to address the missings in the polychoric correlation matrix? I calculate them for each variable combination and store the values in a matrix? But then I would still the missing value for those variables where the polychoric correlation is not defined. Do I replace them with 0 or how did you think doing this? Thank you very much!
Best,
Felix
could you perhaps expand on your idea to use the - cmp - command to address the missings in the polychoric correlation matrix? I calculate them for each variable combination and store the values in a matrix? But then I would still the missing value for those variables where the polychoric correlation is not defined. Do I replace them with 0 or how did you think doing this? Thank you very much!
Best,
Felix
Comment