How to mange names in output text file in R? - r

Question

data:
sam:
res = 0.25 , res1=0.30
bad:
res= 0.30 , res1=0.23
code:
write.table(sam, file = "C:\\Users\\data1.txt", append = F, sep = " ", row.names = TRUE,col.names = TRUE)
write.table(bad, file = "C:\\Users\\data1.txt", append = T, sep = " ",row.names = TRUE, col.names = TRUE)
output of data1:
"x"
"1" 0.25
"x"
"1" 0.3
In fact, I want the output in the text file data1 to be something like this:
res res1
sam 0.25 0.30
bad 0.3 0.23
Any idea is appreciated!

Answer 1

You can supply a named vector in write.table(). You'll just need to set the second set of column names to FALSE since they've already been provided in the first initial call.
sam <- 0.25; bad <- 0.30
write.table(c(sam = sam), col.names = "res", file = "data")
write.table(c(bad = bad), col.names = FALSE, file = "data", append = TRUE)
## read it back in
read.table("data")
# res
# sam 0.25
# bad 0.30
In response to your comment, you can write a helper function to do the appending once the file is initialized. Then we can read it as a list so we have a choice between returning a data frame or a matrix.
sam <- 0.25
bad = 0.3
## initial file creation
write.table(cbind(sam, bad), "data", row.names = FALSE)
## function to append to 'data'
wtFun <- function(x) {
write.table(x, "data", append = TRUE, col.names = FALSE, row.names = FALSE)
}
## new values
sam2 <- 0.99
bad2 <- 25
## append new values
wtFun(cbind(sam2, bad2))
## read the file as a list and set the names
res <- setNames(
scan("data", what = list(0, 0), skip = 1L),
scan("data", what = "", nlines = 1L)
)
## 'res' as a matrix
do.call(rbind, res)
# [,1] [,2]
# sam 0.25 0.99
# bad 0.30 25.00
## 'res' as a data frame
as.data.frame(res)
# sam bad
# 1 0.25 0.3
# 2 0.99 25.0

Answer 2

This will work better if you write sam and bad as data.frames (or matrices), instead of atomics. For example,
sam <- 0.25
bad <- 0.30
##
write.table(
data.frame(res=sam,row.names="sam"),
file="F:/temp/data1.txt",
append=F,sep=" ",
col.names=TRUE,
row.names=TRUE)
##
write.table(
data.frame(res=bad,row.names="bad"),
file="F:/temp/data1.txt",
append=T,sep=" ",
col.names=FALSE,
row.names=TRUE)
##
R> read.table("F:/temp/data1.txt",header=TRUE)
res
sam 0.25
bad 0.30
IMO though, it's not a good idea to force the row.names attribute like this because if you append an object to the file with a row name that already exists, you will get an error when you try to read it back in since the row.names attribute cannot contain duplicate values. You would be better off doing something like
write.table(
data.frame(name="sam",res=sam),
file="F:/temp/data1.txt",
append=F,sep=" ",
col.names=TRUE,
row.names=FALSE)
##
write.table(
data.frame(name="bad",res=bad),
file="F:/temp/data1.txt",
append=T,sep=" ",
col.names=FALSE,
row.names=FALSE)
##
R> read.table("F:/temp/data1.txt",header=TRUE)
name res
1 sam 0.25
2 bad 0.30