[R] Comination of two barcharts and one xyplot

ravi rv15i at yahoo.se
Wed Aug 13 15:07:55 CEST 2008


Hi Rhelpers,
I would like to have some help with a plot which is beyond my capabilities. This plot that I am seeking involves an overlay of two different barcharts and one xyplot.
The code that I have used is the following :
#save(df1,file="M:\\KBR\\df1.RData")
load(file="M:\\KBR\\df1.RData")
# df1$Year.ord created to obtain the right order i.e. 2015M < 2015K
Year.ord<-ordered(Year,levels=c('2003','2005','2007','2009','20011','2013','2015M','2015K'))
# Use reshape package to melt the data frame
library(reshape)
df1m<-melt(df1,id=c("Year","Year.ord"))
library(lattice)
attach(df1m)
barchart(value~Year.ord|variable,scales=list(y="free",x=list(rot=90)),xlab="Year",ylab="No. of Tests *1000",col="blue")
This plot works just fine. But I want to go beyond this. My first data frame (df1) is :
"Year","KI","G48","AvCell","HB","Htens","Impact","Struct","Tens","Year.ord"
"1","2003",15.53,0.3,0.24,37.45,0.76,1.16,3.02,34.05,"2003"
"2","2005",15.64,0.29,0.33,34.64,1.12,1.78,4.2,32.88,"2005"
"3","2007",16.18,0.49,0.59,30.32,1.63,4.23,6.67,30.06,"2007"
"4","2009",17.09,0.67,0.91,29.47,2.27,6.76,9.68,29.25,"2009"
"5","2011",22.39,0.93,1.24,38.03,3.11,9.17,13.18,37.84,"2011"
"6","2013",33.83,1.29,1.87,58.37,4.43,14.06,19.41,57.6,"2013"
"7","2015M",44.91,1.83,2.71,75.54,6.28,20.57,27.51,74.5,"2015M"
"8","2015K",52.22,2.14,3.15,87.71,7.34,23.88,31.98,86.57,"2015K"
My second data frame is (L1) is :
"Year","KIL","G48L","AvCellL","HBL","HtensL","ImpactL","StructL","TensL"
"1",2009,20,1,1,30,2,10,10,35
"2",2011,24,1,1.5,35,3,12,13,38
"3",2013,30,1,2,40,4,14,16,45
What I want, in each panel of the lattice barchart, is to plot histograms of the relevant variable (KI, G48 etc) in one colour for the years 2003 to 2007, and in another colour for the other years. On top of this, I want to have a line plot in each panel with the limits for different years given in the second data frame L1 (as bold lines). 
I would like to have information on the following points :
1. How can I get a combination of these plots in every panel (two histograms and one line plot)? Is it possible?
2. Is it easier to do this with ggplot?
3. I would like to know how I can present the legend also.
Will appreciate any help that I can get.

Thanking You,
Ravi




More information about the R-help mailing list