Double for loop to save several files using R - r

Question

I am trying to do a “for loop” to generate files based on the column "group". I want to create a file for each group. My data is much bigger, but a sample would be:
id = c(1,2,3,4,5,6,7,8,9,10)
group = c(3,1,3,2,1,3,1,2,4,4)
weight = c(10,11,12,13,14,15,16,17,18,19)
index1 = c(50,50,50,50,50,50,50,50,50,50)
index2 = c(50,50,50,50,50,50,50,50,50,50)
data = data.frame(id,group,weight,index1,index2)
for (i in unique(data$group)){
for (j in 1:nrow(data)){
data$weight[j] = ifelse(data$group[j] == data$group[i], 0,data$weight[j])
data$index1[j] = ifelse(data$group[j] == data$group[i], 0,50)
data$index2[j] = ifelse(data$group[j] == data$group[i], 5,50)
}
write.table(data,paste("/home/paulaf/test/",data$group[i],".txt",sep=""),
quote=F,row.names=F,col.names=T)}
It seems to work, but it doesn’t write all the files. Any help would be very much appreciated. Thanks in advance.

Answer 1

Paula,
That code is actually writing four files. But you're overwriting one of those files, so you're only ending up with three.
When you name the file with paste, you're using data$group[i] to generate the name. If you look at those name by using cat() or something similar, you'll notice you have two 3.txt files.
/home/paulaf/test/3.txt
/home/paulaf/test/3.txt
/home/paulaf/test/1.txt
/home/paulaf/test/2.txt
So, that's why your not getting all of you files. Your first 3.txt is overwritten.
Looking a bit more closely at your data object, you can see why this happened.
Your i in your loops is going to have the values 3, 1, 2, and 4. By plugging 1-4 into data$group[i], you're actually pulling out the value of the 1-4th rows in the data$group. Notice that the first and third rows are both group 3.
id group weight index1 index2
1 1 3 0 50 50
2 2 1 0 50 50
3 3 3 0 50 50
4 4 2 0 0 5
5 5 1 0 50 50
6 6 3 0 50 50
7 7 1 0 50 50
8 8 2 0 0 5
9 9 4 18 50 50
10 10 4 19 50 50
Maybe replace your write.table() with this:
write.table(data,paste("/home/paulaf/test/",i,".txt",sep=""),
quote=F,row.names=F,col.names=T)
And one other note to save you future headache: It's often helpful to print some of your variables to the console. It's just a way to get some insight into what's happening.
Also, good luck, keep working with R, you're doing great!

Answer 2

unique(data$group) is a vector of length 4. data$group has a length of 10. You're setting the filenames to the first 4 values of data$group instead of the unique values of data$group.
Try replacing data$group[i] with just i inside the paste that generates the filename, e.g.
for (i in unique(data$group)){
for (j in 1:nrow(data)){
data$weight[j] = ifelse(data$group[j] == data$group[i], 0,data$weight[j])
data$index1[j] = ifelse(data$group[j] == data$group[i], 0,50)
data$index2[j] = ifelse(data$group[j] == data$group[i], 5,50)
}
fileName = paste("/home/paulaf/test/",i,".txt",sep="")
write.table(data,fileName,quote=F,row.names=F,col.names=T)
}

Answer 3

Your problem is very simple. Inside your write.table function, you're pasting the name using data$group[i], but your outside loop is not looping over the indices of the unique groups, but over the group names themselves. Your is are 3 1 2 4, so calling data$group[i] for each of those will result in 3, 3, 1, 2, which means all the filenames are all wrong (one file is replaced and you end up with only 3, for this sample). The solution is then:
write.table(data,paste("/home/paulaf/test/",i,".txt",sep=""),
quote=F,row.names=F,col.names=T)}
It's also slightly more efficiently (and easier to read, imho) to use paste0, so:
write.table(data,paste0("/home/paulaf/test/",i,".txt"),
quote=F,row.names=F,col.names=T)}