How to tag along one more variable to ouput in proc means?

[Phil_NZ]

Hi all SAS Users,

Normally, Output out statement in PROC MEANS only show us the variables called by "BY" statement, and mean, sum,..... called by statement Output out

Can you please tell me how can I tag along another variable call "LOC" to the output here?

So, I have a original dataset called chl_d. The frequency is daily, LOC is 3 characters indexed for a country.

A part is like that

GVKEY	IID	DATADATE	CIK	        CONML	                    LOC     SIC	    ISIN	        SEDOL	EXCHG	AJEXDI	CSHOC	        CSHTRD	    CURCDD	PRCCD	PRCHD	PRCLD	PRCSTD	TPCI	TRFD	ex_rate	gviidkey	prccd_abs	prccd_abs_	    lagprccd_abs_	lagajexdi	lagtrfd	raw_return	  log_raw_return	i	raw_return_abs	dollar_vol	    prccd_	        prchd_	        prcld_	       appday	eta	            c	            eta_lead1	  year	S_2DC
001166	01W	04JAN1999	0000351483	ASM     International NV	NLD  	3559	NL0000334118	5165294	104	    1.0000	33724343.0000	45145.0000	EUR	    4.5000	4.6000	4.4500	10	    0	    1	    0.84753	001166_01W	4.5   	    5.3095465647	.	            .	         .	    .	          .	                85	.	            239699.47966	5.3095465647	5.4275364884	5.2505516029	1	    1.674909242	    1.6695064389	1.673635762	  1999	0.0094466755
001166	01W	05JAN1999	0000351483	ASM     International NV	NLD	    3559	NL0000334118	5165294	104	    1.0000	33724343.0000	32047.0000	EUR	    4.5000	4.6000	4.4500	10	    0	    1	    0.84861	001166_01W	4.5     	5.3027892672	5.3095465647	1	         1	    -0.127266942  .	                85	0.1272669424	169938.48764	5.3027892672	5.4206290286	5.2438693864	1	    1.673635762	    1.6682329589	1.7264615191  1999	0.0354737899
001166	01W	06JAN1999	0000351483	ASM     International NV	NLD	    3559	NL0000334118	5165294	104	    1.0000	33724343.0000	550752.0000	EUR	    4.7000	5.0000	4.6000	10	    0	    1	    0.85324	001166_01W	4.7	        5.5084149829	5.3027892672	1	         1	    3.8776897473  1.3552395503	    85	3.8776897473	3033770.5687	5.5084149829	5.8600159392	5.3912146641	1	    1.7264615191	1.7062769199	1.7121731718  1999	0.02181866
001166	01W	07JAN1999	0000351483	ASM     International NV	NLD  	3559	NL0000334118	5165294	104	    1.0000	33724343.0000	323750.0000	EUR	    4.8500	4.9000	4.6000	10	    0	    1	    0.85682	001166_01W	4.85	    5.660465442	    5.5084149829	1	         1	    2.7603305046  1.0153504206	    85	2.7603305046	1832575.6868	5.660465442	    5.7188207558	5.3686888728	1	    1.7121731718	1.7335061224	1.815432897	  1999	0

I use the code below to get the yearly results, however, the results chl_2d_year just contain some variables, I do not know how to output the variable LOC in chl_2d_year.

/*Calculating CHL for each firm per year*/
proc means data=work.chl_d noprint;
	by gviidkey year;
	var S_2DC appday;
	Output out=chl_2d_year (keep=gviidkey year CHL_2d_c appdaysum where=(appdaysum >=12))
	mean=CHL_2d_c sum=SC APPDAYSUM;
	/*Explain: mean of all S_2DC of one company in a year is called CHL
	sum SC APPDAYSUM means SC is sum of S_2DC and APPDAYSUM is sum of APPDAY*/
run; 

The result is like that

gviidkey	year	CHL_2d_c	    APPDAYSUM
001166_01W	1999	0.0141318615	253
001166_01W	2000	0.0175846612	249
001166_01W	2001	0.0167945019	252
001166_02W	1999	0.0172477814	105
001166_02W	2000	0.0311273775	143
001166_02W	2001	0.0291985294	84
001855_03W	1999	0.0434634019	86
001855_03W	2000	0.0578536264	42
001855_03W	2001	0.1417332222	24
001932_01W	1999	0.0125657319	252

What I want is to insert the column LOC from chl_d to chl_2d_year. Could you please hint me how to do that?

Warmest regards.

[Kurt_Bremser]

You need to merge it in in a separate step. As mentioned in your other thread, SAS procedures won't do anything that is not in their predefined capabilities.

[FreelanceReinh · Accepted solution]

Hi @Phil_NZ,

I think you can use the ID statement (I would like to include a link to the documentation, but the website seems to be down):

id LOC;

Of course, you also need to add LOC to the KEEP= dataset option.

Edit: Here's an excerpt from the offline help (note that "maximum" refers to alphabetical order in case of character variables such as LOC):

Syntax

ID variable(s);

Required Argument

variable(s)

identifies one or more variables from the input data set whose maximum values for groups of observations PROC MEANS includes in the output data set.