I got a data frame with 627 observations of 11 variables, all of them numeric (whit distinct ranges). I try to apply to it a factorial analysis with the factanal
function, but occasionaly I receive the "unable to optimize from this starting value" error message. See, for instance, the following situations:
Try 3 or more factors, lower = 0.005
Error message
Try 6 factors, lower = 0.006
No error message. P-Value = 0.02.
Try 3 or 4 factor, lower = 0.006
Error message
If I, for example, fix the number of factors at five, and start putting very low values fo the lower
paramether, it will work for some particular values, but not for others (no patterns followed). I think I'll go ahead with 6
factors, but can somebody explain why this behavior?