Improving script_001: “Monitor”
[This article was first published on NIR-Quimiometría, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
After having a look to this video: http://www.screenr.com/UxH8 from rtwotutorials, and reading some tutorials, I decided to modified the script from the previous post: Practicing Script with “ R”: Monitor ,
in order to make it more robust .
If there are NA values in our X and Y variables, the results for all the statistics will be NA (see also the video: http://www.screenr.com/loS8), that is not nice, so it´s better to write a warning and stop the analysis to check our data set.
So the new script is:
monitor3<-function(x,y){x1<-is.na(x)
y1<-is.na(y)
if(mean(x1|y1)>0){
print(“There are NA values in X or Y, remove these samples for calculation”)
}else{
n<-length(y)
res<-y-x
par(mfrow=c(2,2))
hist(res,col=”blue”)
plot(x~y,xlab=”predicted”,ylab=”reference”)
abline(0,1,col=”blue”)
l<-seq(1:n)
plot(res~l,col=2)
abline(h=0,col=”blue”)
boxplot(x,y,col=”green”)
{rmsep<-sqrt(sum((y-x)^2)/n)
cat(“RMSEP:”,rmsep,”\n”)}
{(bias<-mean(res))
cat(“Bias :”,bias,”\n”)}
{sep<-sd(res)
cat(“SEP :”,sep,”\n”)}
{r<-cor(x,y)
cat(“Corr :”,r,”\n”)}
{rsq<-(r^2)
cat(“RSQ :”,rsq,”\n”)}
}
}
After having some samples with NA values (not reference value for that sample or not predicted value) the output will be:
“There are NA values in X or Y, remove these samples for calculation”If not it will continue with the calculations:
RMSEP: 0.4373108
Bias : 0.03814815
SEP : 0.4439425
Corr : 0.4864254
RSQ : 0.2366097
Possible improvements I think right now: Bias : 0.03814815
SEP : 0.4439425
Corr : 0.4864254
RSQ : 0.2366097
The function gives me the position of the samples with NA values, so location is easier
or
that the function removes these samples and the calculations can be done without them.
To leave a comment for the author, please follow the link and comment on their blog: NIR-Quimiometría.
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.