search

LSC-MEM-course / random-effect-simulation-assignment-1-rebeccabieber

random-effect-simulation-assignment-1-rebeccabieber created by GitHub Classroom
0 stars 0 forks source link

Singular fits in the wild #2

Open rebeccabieber opened 4 years ago

rebeccabieber commented 4 years ago

I'm getting a singular fit when playing with random slopes for my data (which I'm sure I'm not handling properly anyway, since I'm using percent correct as my DV).

Is it worth looking into that at this point, or not until I'm eventually running it properly (a glmer with binomial distribution and weighting - I think - though right now all that gets me is convergence issues)?

scottrjackson commented 4 years ago

singular fit is not necessarily a convergence issue. I'll pull and see what you're talking about

scottrjackson commented 4 years ago

oh, your data is not in the repo 😄 Is this that data you can't share that you mentioned? If so, maybe paste in some of the output, and tell me where you're getting singularity warnings. But we'll start discussing some convergence issues today, and I'll mention the difference between convergence and singularity.

rebeccabieber commented 4 years ago

Nope, I can share it! Just moved it into that repo. I'm getting both convergence and singular fit issues, depending on which random terms I include

On Tue, Dec 3, 2019, 6:20 PM scottrjackson notifications@github.com wrote:

oh, your data is not in the repo 😄 Is this that data you can't share that you mentioned? If so, maybe paste in some of the output, and tell me where you're getting singularity warnings. But we'll start discussing some convergence issues today, and I'll mention the difference between convergence and singularity.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/LSC-MEM-course/random-effect-simulation-assignment-1-rebeccabieber/issues/2?email_source=notifications&email_token=ALCNCAWTMHYMHBNLSGVAVNLQW3SUVA5CNFSM4JU7MRE2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEF3FBUA#issuecomment-561402064, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALCNCAXMJLRPRMBNQ53MVHTQW3SUVANCNFSM4JU7MREQ .

rebeccabieber commented 4 years ago

Nope, I can share it! Just moved it into that repo. I'm getting both convergence and singular fit issues, depending on which random terms I include