[R] generating a bar chart with two axis for co-linear variable

arun smartpink111 at yahoo.com
Thu Jun 6 20:44:27 CEST 2013


May be this helps:
dat1<- read.table("sampledata.txt",header=TRUE,sep=",",stringsAsFactors=FALSE)
 lst1<-lapply(seq_len(ncol(dat1)),function(i) {Ctdat<- table(dat1[,i]);Ctdat1<-(Ctdat/sum(Ctdat))*100;barplot(Ctdat1,ylim=c(0,100),xlab=colnames(dat1)[i],ylab="Relative Frequency",main=paste("Barplot:",colnames(dat1)[i],sep=" "))})

----- Original Message -----
From: Sudha Krishnan <Sudha.Krishnan at marlabs.com>
To: "r-help at r-project.org" <r-help at r-project.org>
Sent: Thursday, June 6, 2013 2:37 AM
Subject: [R]  generating a bar chart with two axis for co-linear variable

Hello Dimitris,

I was goggling for some help on Sensitivity vs 1-specificity and saw your link.

I hope you can be of help to me in one of the issue that I am facing in generating combo chart(bar chart and plot). I am a novice and have some difficulty in getting this logic correct.

I am give a dataset (I am attaching a sample dataset).

I am using a barplot() and passing values for percentage frequency and the corresponding variables. I am struck here, what my function does is only calculate the frequency for the listed variables and not the frequency percentage. Is there a method or a script with which I can pass the frequency percent and the related values as category columns for x axis?

I will attach the graphs that I have generated so that you can suggest the better way.

Sampledata - Sampledata.txt

What my function does to calculate the frequency with category names in X axis - 1.png

My requirement is to generate percentage frequency of the variable in y1 and not the frequency itself. 2.png (where x categories are missing)


Sudha Krishnan

R-help at r-project.org mailing list
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

More information about the R-help mailing list