I corrected some of my syntax and changed the the glmStepAIC for glm and ti
worked. I tried my command lines with the mdrr dataset (after doing the first
cleaning steps as referred in the manual), the only difference is that I only
took first two columns because it took too much time with one variable and I
got the same resutls: "undefined columns selected". Here are the command lines:
fitControlAUC <- trainControl(method = "cv", number = 10, classProbs = TRUE,
summaryFunction = twoClassSummary, returnResamp = "all", verboseIter = FALSE)
glmROCLum1 <- train(trainDescr[,(1:4)], trainMDRR, method = "glmStepAIC",
tuneLength = 4, metric = "ROC", trControl = fitControlAUC)
So I used glmStepAIC to find the right model and the to obtain the ROC results
I used the glm method using the variables select by glmStepAIC.
Thank you very much for your package help and interest.
Best regards,
J Toledo
Here is my previous session Info
R version 2.13.0 (2011-04-13)Platform: i386-pc-mingw32/i386 (32-bit)
locale:[1] LC_COLLATE=Spanish_United States.1252 LC_CTYPE=Spanish_United
States.1252 [3] LC_MONETARY=Spanish_United States.1252 LC_NUMERIC=C
[5] LC_TIME=Spanish_United States.1252
attached base packages:[1] splines tcltk stats graphics grDevices
utils datasets methods base
other attached packages: [1] e1071_1.5-26 class_7.3-3 caret_4.98
cluster_1.14.0 reshape_0.8.4 plyr_1.5.2 [7] lattice_0.19-26 Rcmdr_1.6-4
car_2.0-10 foreign_0.8-44 survival_2.36-9 nnet_7.3-1 [13]
MASS_7.3-13
loaded via a namespace (and not attached):[1] grid_2.13.0
> Date: Sun, 28 Aug 2011 13:23:24 -0400
> Subject: Re: [R] Trying to extract probabilities in CARET (caret) package
> with a glmStepAIC model
> From: mxk...@gmail.com
> To: tintin...@hotmail.com
> CC: r-help@r-project.org
>
> Can you provide a reproducible example and the results of
> sessionInfo()? What are the levels of your classes?
>
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
