I have used the following code to obtain mean decrease in accuracy for random forest
AAA<-randomForest(CPercentage~., data=data, importance= T)
BBB<-as.data.frame(importance(AAA))
I have created the following dataframe by the above process
%IncMSE IncNodePurity
Campaigntype 3.4815273 207.5336052
Email -1.1606079 2042.5660103
get 4.9073550 35.1237017
free 2.8777972 14.5362957
new 8.4464445 93.3491610
buy 5.9636483 23.9926669
just 4.1262164 21.5611278
month 4.0817729 16.6345631
I am able to obtain the second and third column by BBB$%IncMSE and BBB$IncNodePurity. I want to subset thsi based on the first column which appears unnamed. I am unable to do this. When writing this dataframe to a csv file, it works and all three columns are listed separately. However, I am unable to separate the first two columns. Is there any way i can do this and rename the first column. Will be grateful to anyone who helps
It seems that your "first column" may actually be the index. You can create a column in your data frame from the index, and then reset your index so they're row numbers instead of names.
Try this:
BBB$col_one <- rownames(BBB)
rownames(BBB) <- 1:nrow(rownames)