Dear arun,


> dat1<-read.table("EL.csv",header=TRUE, sep=",",na.strings="NA")


library(reshape2)

res<-lapply(split(dat1,dat1$st),function(x) 
dcast(x,month~year,mean,value.var="discharge"))

> res

$EE
   month      2005      2006      2008      2009
1      1 1.7360776 0.8095275 1.6369044 0.8195241
2      2 0.6962079 3.8510720 0.4319758 2.3304495
3      3 1.0423625 2.7687266 0.2904245 0.7015527
4      4 2.4158326 1.2315324 1.4287387 1.5701019

$WW
   month       2008      2009      2010
1      1  1.4737028  2.314878  2.672661
2      2  1.6700918  2.609722  2.112421
3      3  3.2387775  7.305766  6.939536
4      4  6.7063592 18.745256 13.278218

i then eliminated the first column as i was interested in knowing the row wise mean of each sublist, by using


res1 <- lapply(res, function(x)x[,1 ])

$EE
        2005      2006      2008      2009
1  1.7360776 0.8095275 1.6369044 0.8195241
2  0.6962079 3.8510720 0.4319758 2.3304495
3  1.0423625 2.7687266 0.2904245 0.7015527
4  2.4158326 1.2315324 1.4287387 1.5701019

$WW
         2008      2009      2010
1   1.4737028  2.314878  2.672661
2   1.6700918  2.609722  2.112421
3   3.2387775  7.305766  6.939536
4   6.7063592 18.745256 13.278218


now when i applied the "colMeans" command i was not able to calculate the mean as i got the following error

>'x' must be an array of at least two dimensions

i cant spot the mistake in the last 5 hours.

you help is needed

eliza