Closed isabelslurink closed 2 years ago
It looks like an error with a c ++ library that is used internally in BGGM. So, unfortunately, nothing I can do.
If you cannot figure out a way to use a different computer, you could consider doing a meta-analysis on the subsets that you are able to fit..
I work with R via a Citrix Workspace with limited memory. I want to use BGGM, type = “mixed”. I have the Github version of BGGM in the workspace. When using a data.frame with both int and numeric values, I get the following error: If I convert all variables into integers, the analysis works, but only with a small subsample (eg 500 participants). The BGGM fit object is 40MB with this sample. However, my dataset has 70000 participants, and ~50 variables. I get this error with a larger sample (5000): And with all vars as integers: The model does work on a different dataset outside the workspace on my own PC, but I'm not allowed to transfer the dataset to outside the workspace.
Do you have a sollution for this? Thanks in advance!