Memory Error when sampling my Hierarchical logistic model

Questions
3

Hi, Oriol

Thank you very much for the reply. However, even when I successfully load the model (see below), I still cannot run the

image
image805×486 37.2 KB
sampling even with one chain. The problem seem to be excessive use of RAM to store the model. It takes almost 27 GB of my RAM (I have 32GB of RAM install on the desktop), though the dataset is only 17MB.

Would you suggest a better model specification that allows me to build this hierarchical logistic model?

To be noted is that since each one of my 7420 layer represent a person, there are around 22-ish variables are associated with each person, meaning despite the data might be over 70 thousands rows, those variable only have at most 7420 unique combinations. So, in this case, would you think it might be a good idea to find some way to specifying the those person-related variables in 7420 indexes first, then add them to the rest of the data?

Thank you again for the time reading this!