FarrellDay / miceRanger

miceRanger: Fast Imputation with Random Forests in R
Other
67 stars 12 forks source link

Evaluation failed with error in MiceRanger #22

Open bluepassereaux opened 1 year ago

bluepassereaux commented 1 year ago

I am quite new in Gihub, despite knowing it for years and being such a help. So I will try to give you a proper example as I can.

library(miceRanger)
library(doParallel)
set.seed(1)
# Set up back ends.
cl <- makeCluster(3)
registerDoParallel(cl)

# Perform mice
parTime <- system.time(
  miceObjPar <- miceRanger(
      DATAFRAME,
      valueSelector = "meanMatch",
     m=4,
     maxiter = 6,
     num.threads = 7,
     num.trees = 240,
     parallel = TRUE,
     verbose = FALSE
  )
)
stopCluster(cl)
registerDoSEQ()

Error in miceRanger(RAY, valueSelector = "meanMatch", m = 4, maxiter = 6, : Evaluation failed with error <Error in sample.int(length(x), size, replace, prob): primer argumento inválido

. This is probably our fault - please open an issue at https://github.com/FarrellDay/miceRanger/issues with a reproduceable example. Timing stopped at: 0.06 0 2.27

How can I solve this? How can I search for the "first invalid argument"?

Thank you!!

AnotherSamWilson commented 1 year ago

Does running with parallel = FALSE fix the problem? Also, can you show me a summary of the dataframe, as well as the output of sessionInfo().

bluepassereaux commented 1 year ago

Thank you very much, I appreciate a lot this community.

Yes, I just run without the parallel function and I have the error still. It is very weird because I can run the miceRanger if I move some observations from one table to another. Anyhow, I used to run this table last year and there was no problem.

As well, I tried to verify the data as dataframe or matrix with the correspondent functions and did not work. Anyway my data is a dataframe.

Is there anything I should be looking in the summary?

I cannot give the summary by now as the date is from someone else, but here is my sessionInfo: Screenshot_2