Recommendations for a package capable of doing latent class regression with mixed indicator types

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Recommendations for a package capable of doing latent class regression with mixed indicator types

Weiwen Ng, MPH
Dear R users,

Does anyone have a recommendation for an R package that can:

   1. Handle mixed types of indicators (in my case, both binary and
   Gaussian)
   2. Fit a latent class regression (i.e. use observed covariates to
   predict latent class membership)?
   3. Incorporate programs to do things like profile plots of class
   characteristics?

I'm aware of 3 packages that can do some of what I want. The depmixS4
package does 1 and 2, but doesn't appear to have native graphing
facilities. poLCA does 2 and 3, but it only takes binary and ordinal
indicators. flexmix appears to do 1 and 3, but it doesn't appear to perform
latent class regression.

I've done all 3 of the above in Stata 15, and I'm actually much more
proficient in Stata, but I will have some trouble accessing that program
for this project thanks to institutional changes (long story, don't want to
explain, but I could access that software if there's no R alternative). I
would like to know if I've missed any packages in my survey, or if I'm
misreading the capabilities of the packages, or if graphing is easier than
I think it is.

Thanks,

Weiwen Ng, MPH

University of Minnesota

School of Public Health

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.