Art

Thanks!

Guessing rows won't do it, because I don't know what row. I will look into mixed = and that macro

Peter

-----Original Message----- From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Arthur Tabachneck Sent: Sunday, March 27, 2011 2:29 PM To: SAS-L@LISTSERV.UGA.EDU Subject: Re: Reading numeric data from a CSV file as if it was categorical

Peter,

Will the combination of setting guessingrows and mixed=yes work?

If not, you can find a macro at: http://www.nesug.org/proceedings/nesug03/ps/ps019.pdf

HTH, Art ------- On Sun, 27 Mar 2011 14:13:39 -0400, Peter Flom <peterflomconsulting@MINDSPRING.COM> wrote:

>Hello > > > >I have to read in data from 17 large (each is many megabytes) csv files. >They have some columns where the data is really categorical, but where >the vast majority uses numerical codes, but somewhere (hundreds of >lines in) there is text data. > > > >Of course, this makes SAS complain > > > >Ordinarily, I'd fix all this up in Excel. But with so many large files that >would be a major PITA. So, how can I tell SAS that some columns should >be categorical > > > >I wrote a mini-macro to read in the files: > > > >%macro imp(filename); > >PROC IMPORT OUT= WORK.&filename > > DATAFILE= "C:\personal\Consults\Olivia Dixon\&filename..csv" > > DBMS=CSV REPLACE; > > GETNAMES=YES; > > DATAROW=2; > >RUN; > >%mend imp; > > > >%imp(file1); > > > > > >And I then use it on 17 files. > > > >TIA as always > > > >Peter Flom > >Peter Flom Consulting > >http://www.statisticalanalysisconsulting.com/ > >http://www.IAmLearningDisabled.com