原始数据集

SUBJID	AGERP	sex	arm
01-S001	30 to 39 years	Female	1
01-S003	40 to 49 years	Female	1
01-S005	60 years	Male	1
01-S007	60 years	Male	1
01-S009	40 to 49 years	Female	1
01-S011	60 years	Female	1
01-S013	60 years	Male	1
02-S001	40 to 49 years	Female	1
02-S003	50 to 59 years	Female	1
02-S005	50 to 59 years	Female	1
03-S002	60 years	Female	1
03-S004	40 to 49 years	Female	1
03-S006	50 to 59 years	Female	1
03-S008	30 to 39 years	Female	1
03-S010	50 to 59 years	Female	1
03-S012	60 years	Female	1
03-S014	50 to 59 years	Male	1
03-S016	50 to 59 years	Female	1
03-S018	30 to 39 years	Male	1
03-S020	40 to 49 years	Male	1
01-S002	60 years	Female	2
01-S004	60 years	Male	2
01-S006	60 years	Female	2
01-S008	40 to 49 years	Male	2
01-S010	50 to 59 years	Male	2
01-S012	40 to 49 years	Female	2
01-S014	50 to 59 years	Female	2
02-S002	60 years	Male	2
02-S004	60 years	Male	2
03-S001	50 to 59 years	Female	2
03-S003	40 to 49 years	Male	2
03-S005	60 years	Male	2
03-S007	50 to 59 years	Female	2
03-S009	40 to 49 years	Male	2
03-S011	50 to 59 years	Female	2
03-S013	40 to 49 years	Female	2
03-S015	40 to 49 years	Female	2
03-S017	30 to 39 years	Female	2
03-S019	40 to 49 years	Male	2
03-S021	60 years	Female	2

想通过sql建立一个table变成以下结果

age/sex	group	count
30 to 39 years	1	3
40 to 49 years	1	5
50 to 59 years	1	6
60 years	1	6
30 to 39 years	2	1
40 to 49 years	2	7
50 to 59 years	2	5
60 years	2	7
Male	1	6
Female	1	14
Male	2	9
Female	2	11

我的想法是：
proc sql;
        create table age_group as select arm,count(select AGERP,arm,count(SUBJID) from origin_1 group by AGERP,arm),count(select SEX,arm,count(SUBJID) from origin_1 group by SEX,arm) from origin_1 group by arm;
quit;

但是会报错。
我知道可以分别对AGERP和sex建立两个table，然后set到一起，但是我想知道能不能一次性做到，毕竟我觉得sql很强大！