Lavaan single group cfa. Empty categories but no grouping chosen?

114 Views Asked by At

I am trying to do a single factor cfa with lavaan. It works for every other variable but lavaan seems to have an issue with this specific indicator of this latent variable called "PFC46":

fit <- cfa(model, data=data, ordered=c("other observed variables omitted ","PFC46"))

I then get this error:

Error in lav_samplestats_step1(Y = Data, wt = wt, ov.names = ov.names,  : 
  lavaan ERROR: some categories of variable `PFC46' are empty in group 1; frequencies are [7 15 72 820 0]

But when I check if there are any empty variables I get this result with table()

1   2   3   4   5 
 
17  8  15  75 838 

Does anyone have an idea why its not working? To me it looks like it executes the code as if I added some grouping, which to my knowledge I have not.

Tried to remove any missing values. The column of PFC46 looks exactly like any other where I do not get an error and If i omit PFC46 completely from the analysis it works. Also changed it to a factor, double, integer its all the same. Also double and triple checked the order and the model.

0

There are 0 best solutions below