Dear all,
Is it possible to use chisq.test(x,p) (where x = c(N1,...,Nk) are the
counts in the k "bins" and p is a vector of the probabilities that a
random observation will fall in the various bins) to test the goodness
of fit of a certain distribution for which we had to estimate the
parameters? Is there another function to do that?
Also, how can I use chisq.test or other function to perform a goodness
of fit test for 2 independent samples (to check if they come from the
same distribution)?
Ks.test is not adequate for discrete samples.
And, is it possible to use ks.test to test the goodness of fit of a
certain given distribution that is not in its list (as it is pnorm,
pgamma, ...)?
Again I can't estimate any distribution parameter!
Is it possible to do a Lilliefors test in R?
I couldn't find any of these answers in the literature. Can anyone help
me?
Alexandra
[[alternative HTML version deleted]]
