dear all
attached you will find a csv datasets, there are many steps before these
they work properly. but I have errors in these steps I guess. any help
appreciated.
Step1: convert the data from continuous to categorical
##nautodata is the normalized data. I did it in the previous steps.
MPGCat= c(0,10,15,20,25,30, 35, 40)
MPG <- cut(nautodata$mydata.MPG, MPGCat,labels = c(1:7))
nautodata = data.frame(MPG, nautodata[2:7])
nautodata
Step 2: divided into 10 folds: as follow
fold1= nautodata[1:39,]
fold2= nautodata[40:79,]
fold3= nautodata[80:119,]
fold4= nautodata[120:159,]
fold5= nautodata[160:199,]
fold6= nautodata[200:139,]
fold7= nautodata[240:279,]
fold8= nautodata[280:319,]
fold9= nautodata[320:359,]
fold10= nautodata[360:398,]
datafolds= list(fold1, fold2, fold3, fold4,
fold5,fold6,fold7,fold8,fold9,fold10)
step3:
##conduct 10-fold cross validation on KNN
KNNFoldError= c(0,0,0,0,0,0,0,0,0,0)
MGFoldError= c(0,0,0,0,0,0,0,0,0,0)
for (i in 1:10)
{
trainData = NULL
for(j in 1:10)
{
if(i !=j)
{
trainData = rbind(trainData, datafolds[[j]])
}
else
testData = datafolds[[j]]
}
#print (trainData)
#print(testData)
targetData = trainData$MPG
testTargetData = testData$MPG
trainData$MPG= NULL
testData$MPG = NULL
M1 = knn(train=trainData, test=testData, cl=targetData, k=20)
M2 = MajorityGuessing(testData,MPGCat)
print(table(testTargetData,M1))
print(testTargetData)
print(M1)
print(M2)
KNNFoldError[i] = round(mean(testTargetData != M1), 3)
MGFoldError[i] = round(mean(testTargetData != M2), 3)
print(KNNFoldError)
print(MGFoldError)
}
## these are the error I got:
Quitting from lines 80-86 (l...@m.rmd)
Error in cut.default(nautodata$mydata.MPG, MPGCat, labels = c(1:7)) :
'x' must be numeric
Calls: <Anonymous> ... withCallingHandlers -> withVisible -> eval -> eval
-> cut -> cut.default
Execution halted
______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
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.
