m2 <- dataPCLD1 %>%
impute()%>%
scale()%>%
estimate_profiles(n_profiles = 2,
variances = "equal", covariances = "zero")
So in calculating Latent Profile Analysis there are different results in MPlus and in R (tidyLPA) with the same data. For 2 class solutions, there are different amount of subjects per class (e.g. in R: 30/ 51; Mplus: 37/44) and also different AIC/BIC values. I also tried calculating with the package mclust. Still I get the same results as with tidyLPA.
Does anyone know where the problem relies on?
Kind regards.