Your main problem is it sounds like col_1-3 is not the right name for your format...

-Joe

On Wed, May 4, 2011 at 12:14 PM, oslo <hokut1@yahoo.com> wrote:

> Dear Joe; > > > > Here is the log file. There are a couple error messages. Any idea? > > > > Regads, > > > > Oslo > > > > > > 7421 data ped; > > 7422 input animal_id $ sire_id $ dam_id $; > > 7423 cards; > > NOTE: The data set WORK.PED has 7386 observations and 3 variables. > > NOTE: DATA statement used (Total process time): > > real time 0.04 seconds > > cpu time 0.04 seconds > > > > 14810 ; > > 14811 data for_format; > > 14812 set ped; > > 14813 start=col_1; > > 14814 output; > > 14815 start=col_2; > > 14816 output; > > 14817 start=col_3; > > 14818 output; > > 14819 keep start; > > 14820 run; > > NOTE: Variable col_1 is uninitialized. > > NOTE: Variable col_2 is uninitialized. > > NOTE: Variable col_3 is uninitialized. > > NOTE: There were 7386 observations read from the data set WORK.PED. > > NOTE: The data set WORK.FOR_FORMAT has 22158 observations and 1 variables. > > NOTE: DATA statement used (Total process time): > > real time 0.01 seconds > > cpu time 0.01 seconds > > > > 14821 > > 14822 proc sort nodupkey data=for_format; > > 14823 by start; > > 14824 run; > > NOTE: There were 22158 observations read from the data set WORK.FOR_FORMAT. > > NOTE: 22157 observations with duplicate key values were deleted. > > NOTE: The data set WORK.FOR_FORMAT has 1 observations and 1 variables. > > NOTE: PROCEDURE SORT used (Total process time): > > real time 0.03 seconds > > cpu time 0.03 seconds > > > > 14825 > > 14826 data for_format_fnl; > > 14827 set for_format; > > 14828 label=put(_n_,$6.); *replace 6 with a smaller number appropriate to > your data; > > WARNING: Variable _N_ has already been defined as numeric. > > 14829 fmtname='$RECODEF'; > > 14830 run; > > NOTE: There were 1 observations read from the data set WORK.FOR_FORMAT. > > NOTE: The data set WORK.FOR_FORMAT_FNL has 1 observations and 3 variables. > > NOTE: DATA statement used (Total process time): > > real time 0.00 seconds > > cpu time 0.00 seconds > > > > 14831 > > 14832 proc format cntlin=for_format_fnl lib=work; > > NOTE: Format $RECODEF is already on the library. > > NOTE: Format $RECODEF has been output. > > 14833 quit; > > NOTE: There were 1 observations read from the data set WORK.FOR_FORMAT_FNL. > > NOTE: PROCEDURE FORMAT used (Total process time): > > real time 0.01 seconds > > cpu time 0.01 seconds > > > > 14834 > > 14835 data want; > > 14836 set ped; > > 14837 newcol_1 = put(col_1,$RECODEF.); > > --------- > > 48 > > WARNING: Variable col_1 has already been defined as numeric. > > 14838 newcol_2 = put(col_2,$RECODEF.); > > --------- > > 48 > > WARNING: Variable col_2 has already been defined as numeric. > > 14839 newcol_3 = put(col_3,$RECODEF.); > > --------- > > 48 > > WARNING: Variable col_3 has already been defined as numeric. > > ERROR 48-59: The format RECODEF was not found or could not be loaded. > > 14840 run; > > NOTE: The SAS System stopped processing this step because of errors. > > WARNING: The data set WORK.WANT may be incomplete. When this step was > stopped there were 0 > > observations and 9 variables. > > NOTE: DATA statement used (Total process time): > > real time 0.01 seconds > > cpu time 0.01 seconds > > > ------------------------------ > *From:* Joe Matise <snoopy369@gmail.com> > *To:* oslo <hokut1@yahoo.com> > > *Cc:* SAS-L@listserv.uga.edu > *Sent:* Wed, May 4, 2011 11:59:03 AM > *Subject:* Re: column recoding > > Right, I was explaining a way to do that... something along these lines. > > data for_format; > set have; > start=col_1; > output; > start=col_2; > output; > start=col_3; > output; > keep start; > run; > > proc sort nodupkey data=for_format; > by start; > run; > > data for_format_fnl; > set for_format; > label=put(_n_,$6.); *replace 6 with a smaller number appropriate to your > data; > fmtname='$RECODEF'; > run; > > proc format cntlin=for_format_fnl lib=work; > quit; > > data want; > set have; > newcol_1 = put(col_1,$RECODEF.); > newcol_2 = put(col_2,$RECODEF.); > newcol_3 = put(col_3,$RECODEF.); > run; > > Does that do about what you want? > > -Joe > > On Wed, May 4, 2011 at 11:53 AM, oslo <hokut1@yahoo.com> wrote: > >> Joe; >> >> Thanks for writing. NO. I want creat 3 new columns with recoded or simply >> recode three existing columns. >> >> Oslo >> >> ------------------------------ >> *From:* Joe Matise <snoopy369@GMAIL.COM> >> >> *To:* SAS-L@LISTSERV.UGA.EDU >> *Sent:* Wed, May 4, 2011 11:43:29 AM >> *Subject:* Re: column recoding >> >> So essentially you could simply take all three columns, output them in a >> single column, use one of the various methods available to get unique >> values >> (proc freq, sort nodupkey, etc.), and then process that in a data step >> assigning _n_ as the value for each original value; then use a proc format >> to apply those values. >> >> No? >> >> -Joe >> >> On Wed, May 4, 2011 at 11:40 AM, oslo <hokut1@yahoo.com> wrote: >> >> > Nat; >> > >> > Thanks for writing. No there is no rule. I just gave an examle. Only >> rule >> > is >> > that there are same value in all 3 columns and they are repeating many >> > time. >> > These should be assigned with same value. >> > >> > regards, >> > >> > Oslo >> > >> > >> > >> > >> > ________________________________ >> > From: Nat Wooding <nathani@VERIZON.NET> >> > To: SAS-L@LISTSERV.UGA.EDU >> > Sent: Wed, May 4, 2011 11:31:55 AM >> > Subject: Re: column recoding >> > >> > Oslo >> > >> > Is there some set of rules that determine why IG should be 5 and 14 >> should >> > be 6? >> > >> > Nat Wooding >> > >> > -----Original Message----- >> > From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of >> oslo >> > Sent: Wednesday, May 04, 2011 11:55 AM >> > To: SAS-L@LISTSERV.UGA.EDU >> > Subject: column recoding >> > >> > Dear all; >> > >> > I have very big and messy data set. I need to reorganize it via SAS for >> > analyses >> > the different software (ASREML....). I need to recode the first 3 >> columns >> > of >> > >> > data which they are for pedigree part. Each column contains both numeric >> > and >> > >> > non-numeric info. I want to convert all of them into numeric. Suppose >> the >> > first >> > columns of data is in the following pattern, >> > >> > Your helps are greatly appreciated. Regards, >> > >> > Oslo >> > >> > data I have data I want to have >> > id sire dam id sire dam >> > a4 ch a&M 1 2 3 >> > 9 IG 14 4 5 6 >> > R/h cm R/h 7 8 7 >> > . . . . . . >> > and so on >> > >> > >> > >