Hello ,everyone,
Recently, I have been using extraTrees model in caret package. However, I noticed that the probability function for extraTrees model is set to NULL by using the following scripts:
extratrees_para <- getModelInfo('extraTrees', regex = F)[[1]]
extratrees_para$prob
I noticed that in the original package of extraTress, it can be used to generate probability prediction for classification problems. So I'd like to specify the prob function for extratrees_para.
extratrees_para$prob <- function(modelFit, newdata, submodels = NULL){
as.data.frame(predict(modelFit, newdata, probability = TRUE))
}
extratrees_para$type <- 'Classification'
Then I construct train function to build the model
extratreesGrid <- expand.grid(.mtry=1:2,
.numRandomCuts=1)
modelfit_extratrees <- train(outcome~., data=training_scaled_sel,
method = extratrees_para,
metric = "ROC",
trControl = trainControl(method = 'repeatedcv',
repeats=1,
classProb = T,
summaryFunction = twoClassSummary),
ntree = 3000,
tuneGrid = extratreesGrid)
However, I keep getting this error message which is not very informative
"Error in train.default(x, y, weights = w, ...) :
final tuning parameters could not be determined
In addition: Warning messages:
1: In nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
There were missing values in resampled performance measures.
2: In train.default(x, y, weights = w, ...) :
missing values found in aggregated results"
Below is my session information. I would be really appreciated if anyone can help me with this. Thanks!
sessioninfo()
R version 3.1.2 (2014-10-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
locale:
[1] LC_COLLATE=English_United States.1252
[2] LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C
[5] LC_TIME=English_United States.1252
attached base packages:
[1] grid stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] DMwR_0.4.1 BiocInstaller_1.16.5 caret_6.0-41
[4] ggplot2_1.0.0 lattice_0.20-29 extraTrees_1.0.5
[7] rJava_0.9-6
loaded via a namespace (and not attached):
[1] abind_1.4-3 bitops_1.0-6 BradleyTerry2_1.0-5
[4] brglm_0.5-9 car_2.0-24 caTools_1.17.1
[7] class_7.3-11 codetools_0.2-9 colorspace_1.2-4
[10] compiler_3.1.2 digest_0.6.8 e1071_1.6-4
[13] foreach_1.4.2 gdata_2.16.1 gplots_2.17.0
[16] gtable_0.1.2 gtools_3.4.1 iterators_1.0.7
[19] KernSmooth_2.23-13 lme4_1.1-7 MASS_7.3-35
[22] Matrix_1.1-4 mgcv_1.8-3 minqa_1.2.4
[25] munsell_0.4.2 nlme_3.1-118 nloptr_1.0.4
[28] nnet_7.3-8 parallel_3.1.2 pbkrtest_0.4-2
[31] plyr_1.8.1 pROC_1.8 proto_0.3-10
[34] quantmod_0.4-4 quantreg_5.11 Rcpp_0.11.4
[37] reshape2_1.4.1 ROCR_1.0-7 rpart_4.1-8
[40] scales_0.2.4 SparseM_1.6 splines_3.1.2
[43] stringr_0.6.2 tools_3.1.2 TTR_0.22-0
[46] xts_0.9-7 zoo_1.7-12
