This is not reproducible, and does not look minimal. You'll get better
answers, and probably solve many issues on your own, if you construct
small examples that illustrate the same problem you're having with your
real data.
Addi Wei wrote:
iterations <- 100
nvars <- 4
combined <- rbind(scaleMiceTrain, scaleMiceTest)
reducedSample <- combined
reducedSample <- subset(reducedSample, select = -pID50)
reducedSample <- subset(reducedSample, select = -id)
for (i in 1:iterations)
{
miceSample <- sample(combined[,-c(1,2)],nvars, replace=FALSE)
miceSample$pID50 <- combined$pID50
miceTestSample <- miceSample[47:55,]
miceTrainSample <- miceSample[1:46,]
fit.kknn <- kknn(pID50~., miceTrainSample, miceTestSample)
table(miceTestSample$pID50, fit.kknn$fit)
(fit.train1 <- train.kknn(pID50~., miceTrainSample, kmax=15,
kernel=c("rectangular"), distance=1))
predictedTrain <- predict(fit.train1, miceTrainSample,
miceTrainSample$pID50)
pID50Train <- miceTrainSample$pID50
lmTrain <- lm(predictedTrain~pID50Train)
slm <- summary(lmTrain)
str(slm)
if (i == 1)
{
previousR2 <-slm$r.squared
sink(file="R2outputKKNN.txt", append=TRUE)
previousR2
sink()
}
else if(i!=1)
{
currentR2 <- slm$r.squared
if (previousR2 > currentR2)
{
currentR2 <- previousR2
}
if (previousR2 < currentR2)
{
sink(file="R2outputKKNN.txt", append=TRUE)
currentR2
sink()
}
}
}
In my code above, I can't get sink to work. In summary, I'm trying to write
the first run's R2, which is called "previousR2" to file, and then anytime
"currentR2" > "previousR2", I will write "currentR2" to file. After running
the code above, my file R2outputKKNN.txt is empty...
However, just running the code below writes / works fine:
previousR2 <-slm$r.squared
sink(file="R2outputKKNN.txt", append=TRUE)
previousR2
sink()
______________________________________________
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.
