On 06.11.2012 05:24, Benjamin Caldwell wrote:
Thanks to you all. Modifying the code to use rbind instead of merge worked
like a charm - I can only speculate why.
Because merge is not intended for appending. See its help page.
Uwe Ligges
Best
*Ben Caldwell*
PhD Candidate
University of Cal
Thanks to you all. Modifying the code to use rbind instead of merge worked
like a charm - I can only speculate why.
Best
*Ben Caldwell*
PhD Candidate
University of California, Berkeley
130 Mulford Hall #3114
Berkeley, CA 94720
Office 223 Mulford Hall
(510)859-3358
On Sat, Nov 3, 2012 at 2:19 P
It easier than that. I forgot I can do it entirely within R:
setwd("/temp/csv")
files <- Sys.glob("daily*csv")
output <- file('Rcombined.csv', 'w')
for (i in files){
cat(i, '\n') # write out file processing
input <- readLines(i)
input <- input[-1L] # delete header
writeLines(inp
These are not commands, but programs you can use. Here is a file copy
program in "perl" (I spelt it wrong in the email); This will copy all
the files that have "daily" in their names. It also skips the first
line of each file assuming that it is the header.
perl can be found on most systems.
Jim,
Where can I find documentation of the commands you mention?
Thanks
On Sat, Nov 3, 2012 at 12:15 PM, jim holtman wrote:
> A faster way would be to use something like 'per', 'awk' or 'sed'.
> You can strip off the header line of each CSV (if it has one) and then
> concatenate the files t
A faster way would be to use something like 'per', 'awk' or 'sed'.
You can strip off the header line of each CSV (if it has one) and then
concatenate the files together. This is very efficient use of memory
since you are just reading one file at a time and then writing it out.
Will probably be a
On the absence of any data examples from you per the posting guidelines, I will
refer you to the help files for the melt function in the reshape2 package.
Note that there can be various mixtures of wide versus long... such as a wide
file with one date column and columns representing all stock p
Jeff,
If you're willing to educate, I'd be happy to learn what wide vs long
format means. I'll give rbind a shot in the meantime.
Ben
On Nov 2, 2012 4:31 PM, "Jeff Newmiller" wrote:
> I would first confirm that you need the data in wide format... many
> algorithms are more efficient in long forma
I would first confirm that you need the data in wide format... many algorithms
are more efficient in long format anyway, and rbind is way more efficient than
merge.
If you feel this is not negotiable, you may want to consider sqldf. Yes, you
need to learn a bit of SQL, but it is very well integ
Dear R help;
I'm currently trying to combine a large number (about 30 x 30) of large
.csvs together (each at least 1 records). They are organized by plots,
hence 30 X 30, with each group of csvs in a folder which corresponds to the
plot. The unmerged csvs all have the same number of columns (5)
10 matches
Mail list logo
