topic Re: Attempt at replicating a demographic table. Which PROC is best? Also how to do it? in SAS Programming

Attempt at replicating a demographic table. Which PROC is best? Also how to do it?

Hello_there — Sat, 09 Jan 2021 14:50:52 GMT

Hi,

I'm trying to replicate this table.

PROC TABULATE data=data;
	class TREATMENT1 GENDER;
	table GENDER ALL, TREATMENT1 all;
	KEYLABEL ALL="Total";
	run;

1. I need to have a fourth column that combines High and Low.

2. I also need to incorporate percents in the columns, so it equals N(%) for the categorical variable.

Thanks

Re: Attempt at replicating a demographic table. Which PROC is best? Also how to do it?

PaigeMiller — Sat, 09 Jan 2021 15:45:59 GMT

https://support.sas.com/resources/papers/proceedings/proceedings/sugi27/p202-27.pdf

Re: Attempt at replicating a demographic table. Which PROC is best? Also how to do it?

Reeza — Sat, 09 Jan 2021 16:07:49 GMT

Your best bet:

https://communities.sas.com/t5/SAS-Communities-Library/Demographic-Table-and-Subgroup-Summary-Macro-TABLEN/ta-p/634030

@Hello_there wrote:

Hi,

I'm trying to replicate this table.
PROC TABULATE data=data;
	class TREATMENT1 GENDER;
	table GENDER ALL, TREATMENT1 all;
	KEYLABEL ALL="Total";
	run;
1. I need to have a fourth column that combines High and Low.

2. I also need to incorporate percents in the columns, so it equals N(%) for the categorical variable.

Thanks

Re: Attempt at replicating a demographic table. Which PROC is best? Also how to do it?

Ksharp — Sun, 10 Jan 2021 10:51:00 GMT

PROC SQL. Here is an example.


%macro report_detail(dataset=,var=,time=);
/*For 第一个变量的 统计量*/
proc univariate data=&dataset(where=(time=&time)) outtable=temp_summary noprint;
class drug;
var &var ;
run;

/*For 第一个变量的 Geomean(CV)*/
ods select none;
ods output Equality=Temp_eq ConfLimits=Temp_ge TTests=Temp_tp;
proc ttest data=&dataset(where=(time=&time)) dist=lognormal;
class drug;
var &var;
run;
ods select all;
data Temp_ge_1;
 set Temp_ge(where=(Method is missing));
 drug=input(class, drug_ifmt.);
run;


proc sql;
create table temp_report_%sysfunc(translate(&time,_,.))  as
/*For 第一个变量的 N(Missing)*/
select &time as a,catx(' ',&time,'hr') as b length=10 ,1 as c,'N(Missing)' as d length=20,
 drug as id format=drug_fmt.,
 cats(_NOBS_,'(',_NMISS_,')') as value
 from temp_summary
union
/*For 第一个变量的 Mean(SD)*/
select &time as a,catx(' ',&time,'hr') as b, 2 as c,'Mean(SD)' as d,
 drug as id ,
 cats(put(_MEAN_,10.4),'(',put(_STD_,10.4),')') as value
 from temp_summary
union
/*For 第一个变量的 Median*/
select &time as a,catx(' ',&time,'hr') as b,3 as c,'Median' as d,
 drug as id ,
 put(_MEDIAN_,10.4 -l) as value
 from temp_summary
union
/*For 第一个变量的 Q1,Q3*/
select &time as a,catx(' ',&time,'hr') as b,4 as c,'Q1,Q3' as d,
 drug as id ,
 cats(put(_q1_,10.4),',',put(_q3_,10.4)) as value
 from temp_summary
union
/*For 第一个变量的 Min,Max*/
select &time as a,catx(' ',&time,'hr') as b,5 as c,'Min,Max' as d,
 drug as id ,
 cats(put(_MIN_,10.4),',',put(_MAX_,10.4)) as value
 from temp_summary
union
/*For 第一个变量的 CV*/
select &time as a,catx(' ',&time,'hr') as b,6 as c,'CV' as d,
 drug as id ,
 put(_CV_/100,10.4 -l) as value
 from temp_summary
union
/*For 第一个变量的 Geomean(CV)*/
select &time as a,catx(' ',&time,'hr') as b,7 as c,'Geomean(CV)' as d,
 drug as id , 
 cats(put(GeomMean,10.4),'(',put(CV,10.4),')') as value
 from temp_ge_1
;
quit;
%mend report_detail;

Re: Attempt at replicating a demographic table. Which PROC is best? Also how to do it?

Reeza — Tue, 12 Jan 2021 20:18:28 GMT

A 20 year old paper that's still valid and appropriate...one of the things that's nice about SAS.