来自这样的数据集:
by_var marker
A 0
A 0
A 1
B 0
B 0
B 1
你可以得到一个看起来像这样的数据集(如果每个观察都是一分钟):
by_var marker minute
A 0 1
A 0 2
A 1 3
B 0 1
B 0 2
B 1 3
通过按by_var 排序并使用简单的data 步骤从您的数据集中:
data my_data2;
set my_data;
by by_var;
if first.by_var then minute = 0;
minute+1;
output;
run;
或者您是否需要更改类似的内容(使用时间变量):
by_var marker time
A 0 12:34:01
A 0 12:34:59
A 0 12:35:01
A 0 12:36:12
A 1 12:36:50
B 0 12:34:01
B 0 12:34:09
B 0 12:34:59
B 0 12:36:12
B 1 12:37:50
到这里:
by_var marker time time2 minutes Cumsum_minutes
A 0 12:34:01 . 0 0
A 0 12:34:59 12:34:01 0 0
A 0 12:35:01 12:34:59 1 1
A 0 12:36:12 12:35:01 1 2
A 1 12:36:50 12:36:12 0 2
B 0 12:34:01 12:36:50 0 0
B 0 12:34:09 12:34:01 0 0
B 0 12:34:59 12:34:09 0 0
B 0 12:36:12 12:34:59 2 2
B 1 12:37:50 12:36:12 1 3
代码:
data my_data;
input by_var $ marker time time.;
format time time.;
cards;
A 0 12:34:01
A 0 12:34:59
A 0 12:35:01
A 0 12:36:12
A 1 12:36:50
B 0 12:34:01
B 0 12:34:09
B 0 12:34:59
B 0 12:36:12
B 1 12:37:50
;
run;
proc sort data=my_data;by by_var;run;
data final;
set my_data;
format time2 time.;
by by_var;
time2=lag(time);
if first.by_var then do;
minutes = 0;
Cumsum_minutes = 0;
end;
else minutes=intck("minutes",time2,time);
Cumsum_minutes+minutes;
run;