Replacing missing values ​​with mean groups in R-error: out of bounds

I have a huge file that looks like this:

 V1    SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9
GROUP1   1    NA   2     1    1   NA   1    1    2
GROUP1   1     2  NA     0    0    2   1   1     NA
GROUP1   0     2   2     0   NA    1   1   1     2
GROUP2   1     2   1     1    1   NA   2   0     2
GROUP2   1     1   1    NA    0    1   0   1    NA
GROUP2   1     1  NA     1    0    1   NA  1     0

      

        
        
        
      

    

What I need to do is replace the missing values ​​with the mean. I did it in a small example and it works. However, when I do this on a large file, I get an error: "index out of bounds". What I am doing is: Create a list with groups that I want to keep for further analysis:

group.list = unique(data_file$V1)

      

        
        
        
      

    

Now I need to do the average for each column based on the group:

A<-colMeans(data_file[data_file$V1 == group.list[1],-1],na.rm=T)
for(i in 2:length(group.list)){
  A <- rbind(A, colMeans(data_file[data_file$V1 %in% group.list[i],-1], na.rm=T))
}
rownames(A)<-group.list

      

        
        
        
      

    

There are some column averages (SNPs) that are missing. So I did this:

SNP.present <- which(A[1,]>=0)
for(i in 2:length(group.list)){
  SNP.present <- intersect(SNP.present,which(A[i,]>=0))
}
A <- A[,SNP.present]

data_file1 = data_file[,c(1,SNP.present+1)]
for(i in 1:dim(data_file1)[1]){
  a <- which(is.na(data_file1[i,]))
  if(length(a)>0){ 
    data_file1[i,a]<-A[data_file1$V1[i],a] 
  }
}

      

        
        
        
      

    

When I run this on a small dataset, it works. However, when I run the full dataset, I get the error:

Error in [data_file1 $ V1 [i], a]: index out of bounds

Does anyone know what might be wrong?