search

kollerma / robustlmm

This is an R-package for fitting linear mixed effects models in a robust manner. The method is based on the robustification of the scoring equations and an application of the Design Adaptive Scale approach.
28 stars 9 forks source link

Error in .local(x, y, ...) : negative length vectors are not allowed (Memory Limits) #14

Closed katPa90 closed 3 years ago

katPa90 commented 3 years ago

Hi! I hope this is the right place to ask this question. I am relatively new to using linear models. For one specific question, my supervisor asked me to use rlmer (and this package) to test an effect. However, my dataset is relatively long (76 subjects à 770 trials). I model three fixed and one random effects. However, by using rlmer I reach the capacity limits of my processor, as I receive the error

Error in .local(x, y, ...) : negative length vectors are not allowed

[I do not receive this error if I use only a subset of the subjects and trials.]

Is there a way to get this running on my machine?

Best Kat

kollerma commented 3 years ago

Dear Kat

Have you tried using rlmerRcpp instead of rlmer?

rlmerRcpp uses an alternative implementation that does more in c++. That implementation might be slightly more memory efficient. Otherwise you're going to have to simplify your model or run it on a larger machine.

Best, Manuel