[R-sig-ME] Efficient mixed logistic reg w 500k individuals

[Phillip Alday]

If you're up for branching out from R a little, then MixedModels.jl in
Julia should be able to handle this. And if it can't, then I would very
interested in fixing that shortcoming. :)  I'm also willing to help you
get the Julia code running.

The previous paragraph was a bit of self advertising, but I can add even
more in. :D With my JellyMe4 package, you can fit a model in Julia, then
move it back to lme4/R to take advantage of the wonderful ecosystem for
plotting, post-hoc comparisons, etc. that's grown up in R general and
around lme4 in particular.


Best,
Phillip

On 24/12/20 2:19 am, Mitchell Maltenfort wrote:
> Here’s a fun one for you (I hope)
> 
> I’m mucking about with a logistic regression that may have about 30 million
> records for half a million individuals.
> 
> Yes, I have a large RAM machine - 64 Gig.  And I’ve used nAGQ 0 and other
> recommendations from
> http://angrystatistician.blogspot.com/2015/10/mixed-models-in-r-bigger-faster-stronger.html?m=1
>  which should be reasonable for the large data.
> 
> It works but I’d still be interested in tweaks to improve speed or
> accuracy.  Any ideas?
>