getPartitionFromDataframe <- function(dataframe, partitioning, subtract=FALSE, direction="rows") {
	if(direction == "rows"){
		if(subtract==TRUE){
			result <- dataframe[-partitioning,]
		} else {
			result <- dataframe[partitioning,]
		}
	} else {
		if(subtract==TRUE){
			result <- dataframe[,-partitioning]
		} else {
			result <- dataframe[,partitioning]
		}
	}
	result
}

getPartitionFromList <- function(list, partitioning, subtract=FALSE) {
	if(subtract==TRUE){
		result <- list[-partitioning]
	} else {
		result <- list[partitioning]
	}
	result
}

removeNearZeroVariance <- function(dataframe) {
	nzv <- nearZeroVar(dataframe)	
	result <- dataframe[-nzv,]
}

removeCorrelating <- function(dataframe) {
	cor <- findCorrelation(cor(dataframe),cutoff=0.99)	
	result <- dataframe[-cor,]
}

performSVMtraining <- function(trainSet, trainLab) {
	tc <- trainControl(method="LGOCV", p = 0.75, number = 25, returnResamp = "final", verboseIter=TRUE )
	svmTrain <- train(trainSet, trainLab, "svmPoly", tuneLength=3, trControl=tc, importance=TRUE)
	svmTrain
}

performRFtraining <- function(trainSet, trainLab) {
	tc <- trainControl(method="LGOCV", p = 0.75, number = 20, returnResamp = "final", verboseIter=TRUE )
	rfTrain <- train(trainSet, trainLab, "rf", tuneLength = 12, trControl = tc, importance=TRUE)
	rfTrain
}

predictionConfusion <- function(svmTrain, testAtData, testAtFactor){
	predValuesSVM <- extractPrediction(list(svmTrain), testX=testAtData, testY=testAtFactor)
	outcome <- predValuesSVM[predValuesSVM$dataType == "Test", ]
	confusionMatrix( outcome$pred, outcome$obs )
}

getImpVars <- function(svmResult) {
	impV <- varImp( svmResult )$importance
	impV <- impV[order(as.vector(rowSums(impV)),decreasing=TRUE),]
	impV
}