[R] reshape data frame

arun smartpink111 at yahoo.com
Sat Nov 16 00:56:59 CET 2013


var1 <- load("reshape_data.frame.RData")
##It is better not to name the objects with function names.
dat1 <- data
 reshape1 <- reshape
names(dat1)[grep("X\\d+",names(dat1))] <- gsub("[[:alpha:]]","X_",names(dat1)[grep("X\\d+",names(dat1))])
res1 <- reshape(dat1,direction="long",varying=7:ncol(dat1),sep="_")
res2 <- res1[with(res1,order(Yr,Seas,Flt.Svy,Gender,Part,NSAMP)),-9]
row.names(res2) <- 1:nrow(res2)
colnames(res2) <- colnames(reshape1)
#[1] TRUE


Some advice on transforming my data would be appreciated.  Attached is 
an .Rdata image with my examples (reshape_data.frame.Rdata). Within the 
image are 2 objects: 

1) The "data" object contains an example of my original data 
format. The columns "X20" : "X50" are histogram bins, and the rows 
beneath are the number of observations (ie. counts) within those bins. 
The other columns; "Yr", "Seas", "Flt.Svy", "Gender", "Part", "NSAMP" 
are observation associated with each bin count.   

2) The "reshape" object is the format I need to transform the "data" object into. 

I think the reshape() function is designed for this, and my 
"data" object is in wide format, and my "resphape" object would be in 
long format.  If indeed reshape() is the best way to do this, I'm 
looking for help in calling the function to transform my data.   

Many thanks, 


More information about the R-help mailing list