Can't run any model after update

30 views
Skip to first unread message

Lucas Perry

unread,
Jun 14, 2024, 1:06:21 PMJun 14
to Genomic SEM Users

I’m revisiting some data I’d previously run through genomicSEM successfully. I updated the package in order to use the Q-SNP option in userGWAS, and found that I can no longer run any model, using either the userGWAS or usermodel functions. They both produce similar errors, either “object 'ReorderModel' not found” or “object 'fitCFI' not found”, respectively.

I’ve tried a couple things to troubleshoot so far. I’ve confirmed that there are no missing values or mislabelled columns in my LDSC (although I don’t see how there could, be, since like I said the data worked previously). I’ve tried rerunning munge, LDSC, ect. I’ve tried different data. I’ve tried different models, and found the error persists even when specifying a simple common factor model. And naturally I’ve tried just reinstalling genomicSEM. No luck.

Any help would be appreciated.

Thank you,

Lucas

agro...@gmail.com

unread,
Jun 14, 2024, 3:05:03 PMJun 14
to Genomic SEM Users
Hi Lucas, 

Thanks for providing the different details and helpful to know that you've tried different options (including reinstalling) and still getting the same result. It sounds like it doesn't have to do with the new Q_SNP argument since usermodel is returning the same error. There haven't been any major updates to usermodel recently so I'm not sure why it would suddenly stop working. These errors are normally about something in your model syntax not matching up with the naming of the variables in the LDSC object. Could you paste your LDSC column names in the S matrix and the full set of code you ran for usermodel? 

Thanks, 
  Andrew  

agro...@gmail.com

unread,
Jun 14, 2024, 4:37:17 PMJun 14
to Genomic SEM Users

Hi Lucas, 

Turns out it was a lavaan update to how it expects certain arguments to be specified. I've updated the code in various places and did a few test runs and seems to be working now. If you redownload GenomicSEM, restart R, and retry you should be good to go, but let me know if you are still having issues. 

Best, 
  Andrew
Reply all
Reply to author
Forward
0 new messages