search

data-edu / tidyLPA

Easily carry out Latent Profile Analysis (LPA) using open-source or commercial software
https://data-edu.github.io/tidyLPA/
Other
56 stars 17 forks source link

Fix for Mclust models that fail #113

Closed cjvanlissa closed 5 years ago

cjvanlissa commented 5 years ago

Here's a fix for Ricardo's difficulties. Josh, can you test these functions for both Mplus and Mclust?

jrosen48 commented 5 years ago

Yes, will do. Sorry, I didn't see this until now. My GitHub notifications were messed up (but are fixed for subsequent issues/PRs!).

jrosen48 commented 5 years ago

Nice, this worked for me.

Two questions:

  1. Should we print a message that a particular model did not estimate/did not converge/that an error occurred?

  2. At present, when a model is estimated but does not converge, a blank row is printed in the output: thoughts on a) removing the row from the output, b) adding another column with a variable to indicate that an error occurred, or c) adding NA for all of the possible values?

jrosen48 commented 5 years ago

@cjvanlissa tagging you on this re: 2 questions above.

cjvanlissa commented 5 years ago

I think it already does this.. only the print method replaces NA with blank space. The summary function shows an extra column to indicate which models had errors.