[Rd] [R] read.csv behaviour

Jim Lemon jim at bitwrit.com.au
Wed Sep 28 12:17:20 CEST 2011


On 09/28/2011 09:23 AM, Mehmet Suzen wrote:
>
> This might be obvious but I was wondering if anyone knows quick and easy
> way of writing out a CSV file with varying row lengths, ideally an
> initial data read from a CSV file which has the same format. See example
> below.
>
>
> I found it quite strange that R cannot write it in one go, so one must
> append blocks or post-process the file, is this true? (even Ruby can do
> it!!)
>
> Otherwise it puts ,"","" or similar for missing column values in the
> shorter length rows and fill=FALSE option do not work!
>
> I don't want to post-process if possible.
>
> See this post:
> http://r.789695.n4.nabble.com/Re-read-csv-trap-td3301924.html
>
> Example that generated Error!
>
> writeLines(c("A,B,C,D",
>               "1,a,b,c",
>               "2,f,g,c",
>               "3,a,i,j",
>               "4,a,b,c",
>               "5,d,e,f",
>               "6,g,h,i,j,k,l,m,n"),
>             con=file("test.csv"))
>
> read.csv("test.csv")
> try(read.csv("test.csv",fill=FALSE))

Hi Mehmet,
The example doesn't need to call "file", writeLines does it for you. It 
worked for me:

  writeLines(c("A,B,C,D",
               "1,a,b,c",
               "2,f,g,c",
               "3,a,i,j",
               "4,a,b,c",
               "5,d,e,f",
               "6,g,h,i,j,k,l,m,n"),
             con="test.csv")

and to get the original object back, use:

readLines("test.csv")

The reason you can't use read.csv is that it returns a data frame, and 
that object can't have elements of unequal length. If you want an object 
with elements of unequal length, try:

as.list(readLines("test.csv"))

Jim



More information about the R-devel mailing list