Possible to regroup rows under subheader column in PROC REPORT with ASSESS option?

RobertWF1 — Fri, 08 Jul 2022 18:08:15 GMT

I've built a dataset containing row percent and count values from a two-way frequency table + the chi-squared p-value and total sample size.

Can I regroup the row percent and count columns under one of the variables as a subheader?

For example:

proc freq data=sashelp.cars; 	
	tables origin*cylinders / chisq totpct outpct list out=freqpct(keep=cylinders origin count pct_row); 
	output out=chisq(keep=N	P_PCHI) all;
run;
data freqpct;
set freqpct;
	var=1;
run;
data chisq;
set chisq;
	var=1;
run;
proc sql;
	create table chisq_freqpct
	as select *
	from freqpct a
	inner join
	chisq b
	on a.var=b.var;
quit;
proc report data=chisq_freqpct;
  	column origin cylinders count pct_row N P_PCHI;
  	define origin / group;
  	define cylinders / group;
	define count / display;
	define pct_row / display;
	define N / group;
	define P_PCHI / group;
run;

Is there a way to use the ASSESS option to display separate columns of pct_row and count values by each value of origin ("Asia", "Europe", "USA"), but still retain N and P_PCHI in the single top row of the table?

Re: Possible to regroup rows under subheader column in PROC REPORT with ASSESS option?

RobertWF1 — Fri, 08 Jul 2022 19:09:10 GMT

Almost there - if I run the following proc report the header and subheaders are correct, but I'm seeing duplicate cylinder rows in the table:

proc report data=chisq_freqpct;
  	column cylinders origin,(count pct_row) N P_PCHI;
  	define cylinders / display /*group*/;
	define origin / across;
	define count / display;
	define pct_row / display;
	define N / group;
	define P_PCHI / group;
run;

topic Re: Possible to regroup rows under subheader column in PROC REPORT with ASSESS option? in SAS Programming

Possible to regroup rows under subheader column in PROC REPORT with ASSESS option?

Re: Possible to regroup rows under subheader column in PROC REPORT with ASSESS option?