CODEWITHSUNDEEP
rstatistics
IcedCoffee
IcedCoffee

Reputation: 385

Linear Mixed Effects Model not finishing in R

I have a problem with a linear mixed effects model not completing in R.

I use the package lmekin to run the model, as I am fitting a kinship matrix as a random effect. The formula I am using is:

model <- lmekin(disease ~ score + age + (1|IID), data=phenotype_df, varlist=kinship_mx)

phenotype_df is a dataframe, with 12000 individuals, each with a disease status (1 or 0), a score and their age. (Note: I am treating the disease status here as quantitative). kinship_mx is a kinship matrix, ie. a positive-definite matrix (12000x12000) with diagonals ~0.5.

This model runs in around 1 minute when I use ~1000 individuals. However, I left the model to run with all 12000 individuals on our large server and it has not finished after 5 days. Could anyone suggest a reason as to why this is happening and if there are any solutions to enable the model to complete?

Upvotes: 0

Views: 162

Answers (1)

IcedCoffee
IcedCoffee

Reputation: 385

Just to answer this question - they finished after three weeks. It just takes a long time.

Upvotes: 0

Related Questions