[R] Conversion problem with write.csv and as.character()

Sarah Goslee sarah.goslee at gmail.com
Thu Mar 10 16:25:14 CET 2016

Hi Peter,

We really need a reproducible example to solve this kind of question.
Please use dput(head(yourdata)) to provide a sample of data (or make
up fake data that shows the same problems), and provide the code
you're using.


On Thu, Mar 10, 2016 at 6:14 AM, Peter Neumaier
<peter.neumaier at gmail.com> wrote:
> Hi all, sorry for double/cross posting, I have sent an initial, similar
> question
> accidentally to r-sig-finance.
> I am writing a matrix (typeof = double) into a CSV file with write.csv.
> My first column of the matrix is a date in the form yyyy-mm-dd hh:mm:ss:
>> a_fetchdata[1,0]
> 2016-02-09 07:30:00
>> typeof(a_fetchdata[1,0])
> [1] "double"
> My CSV file contains a sequence of integers (from 1 to x) instead of the
> expected date.
> I tried to convert, but ran into "Error in dimnames":
>> as.character(first_fetchdata[1,0])
> Error in dimnames(cd) <- list(as.character(index(x)), colnames(x)) :
>   'dimnames' applied to non-array
> Called from: as.matrix.xts(x)
> Browse[1]> c
> a) How can I prevent the conversion into integers to happen when writing
> into CSV?
> b) if a) is not do-able: how can I convert the date in double format to
> chars (i.e. with as.character) ?
> Thanks in advance,
> Peter

More information about the R-help mailing list