I just wanted to send one final message out about this. My previous post
technically works, but is quite slow with large files/strings. I've
decided instead to do it exactly in the way that I was trying to avoid.
So just to set the record straight, these are the functions I'm going to
be using (the second two only so I can avoid a little extra thinking):
-------
write.csv.str <- function(df, row.names=F) {
filepath <- tempfile()
f <- file(filepath, "w")
write.csv(df, f, row.names=row.names)
close(f)
s <- readfile(filepath)
s
}
read.csv.str <- function(s) {
read.csv(text=s)
}
readfile <- function(filepath) {
text <- readChar(filepath, file.info(filepath)$size)
text
}
-------
Thanks for the help everyone.
Cheers,
Thomas
On 03/12/2015 02:12 PM, John McKown wrote:> On Thu, Mar 12, 2015 at 12:15 PM, Thomas Nyberg <tomuxiong at
gmail.com> wrote:
>> Hello,
>>
>> I've found the following useful functionality:
>>
>>> s <- 'cola,colb\n1,2\n2,3\n'
>>> read.csv(text=s)
>> cola colb
>> 1 1 2
>> 2 2 3
>>
>>
>> But I haven't found a similar option in write.csv. I.e. I would
like to
>> "write" a dataframe to a string. What would be the easiest
way to go
>> about such a thing? Right now I can only think of using a file as an
>> intermediary, but that seems a bit silly. Thanks for any help.
>>
>> Cheers,
>> Thomas Nyberg
>
> Perhaps something like:
>
> textConn=textConnection("textCSV",open='w');
> write.csv(file=textConn,s)
> print(textCSV)
>
>
>