I have a correlation matrix and when I try to run a maximum-likelihood factor analysis on it using the "factanal" function I get the following error:
factanal(covmat=matches_adjusted, factors=5, rotation=none, na.omit(matches_adjusted))
Warning in log(e) : NaNs produced
Error in optim(start, FAfn, FAgr, method = "L-BFGS-B", lower = lower, :
L-BFGS-B needs finite values of 'fn'
What could be causing this?