Error in R (mice package), too many weights

r r-mice
Carlos Pesquera Alonso picture Carlos Pesquera Alonso · Feb 16, 2015 · Viewed 7.3k times · Source

I get the following error while imputing missing cases with the mice function from the library "mice"

 Error in nnet.default(X, Y, w, mask = mask, size = 0, skip = TRUE, softmax = TRUE,  :
too many (1104) weights

The problem is generated by the function mice.impute.polr and mice.impute.polyreg because of the default maximum number of weights. I can not solved it by using the command substitute and neither by copying the functions' code and writing the new functions mice.impute.polr and mice.impute.polyreg (because of a function I cannot find call augment). I've told that I should go to the source code to modify it.

How can I do it? Are there any other solution?

Answer

Graham Parsons picture Graham Parsons · Apr 20, 2015

The neural net function called by mice() is stopping because the "maximum allowable number of weights" has been exceeded. The MaxNWts argument to nnet is there to prevent running code that will take a very long time to complete.

If you don't mind waiting then you can increase the MaxNWts parameter by passing it directly to mice(), which will be picked up by nnet():

mice(data = df_with_nas, MaxNWts = 2000)

Related questions

Imputation MICE in R still NA left in dataset

After running MICE package, the number of missing values are shrinked from 147428 to 46093 in each of the 5 complete imputation sets. But isn't it supposed to be 0 NAs instead??? Thanks! Here is my MICE code: imp = mice(newdata) imputationSet1 = complete(imp) …

Read More
Understanding memory use in R: "cannot allocate vector of size"

I have read about other people's problems with memory use, but none of them help with the following issue. I have a 64 bit R environment with 16GB of RAM. Import .gz file of size 229MB. (the unzipped version is 921MB) …

Read More
Do imputation in R when mice returns error that "system is computationally singular"

I am trying to do imputation to a medium size dataframe (~100,000 rows) where 5 columns out of 30 have NAs (a large proportion, around 60%). I tried mice with the following code: library(mice) data_3 = complete(mice(data_2)) After the first iteration I …

Read More