Thanks very much, Cynthia. I said that "out=options should generate a dataset that have similar data values and structure as the report shows" -- I read this from one SAS report procedure book and I believe it is true in most of the cases. As your example shows, the detail report row should match the corresponding values in the out=option dataset. the differences are mainly those rows generated by compute after/before,break after/before blocks. I use out= option is mainly because I wanna see the innate calculation of proc report and utilize it as a reference aiding me to better use compute block inside proc report. If I exclude summay report rows(with _break_ variable with non-missing values) then the dataset in the out= options should match the value in the report output(let's assume noprint option is not used). let me show you the first page in report here: Retail sales in millions of $ DATE YEAR MONTH DAY haha ------------------------------------------------------------ $220 80Q1 1980 1 1 1 $257 80Q2 1980 4 1 2 $258 80Q3 1980 7 1 3 $295 80Q4 1980 10 1 4 $247 81Q1 1981 1 1 5 $292 81Q2 1981 4 1 6 $286 81Q3 1981 7 1 7 $323 81Q4 1981 10 1 8 and then the second page Retail sales in millions of $ DATE YEAR MONTH DAY haha ------------------------------------------------------------ $284 82Q1 1982 1 1 1 $307 82Q2 1982 4 1 2 $318 82Q3 1982 7 1 3 $343 82Q4 1982 10 1 4 $299 83Q1 1983 1 1 5 $351 83Q2 1983 4 1 6 $359 83Q3 1983 7 1 7 $384 83Q4 1983 10 1 8 there are several more pages.... now you can notice, haha can show the row number in the current page. so I expect in out= dataset ,haha should be in 1,2,3..8,1,2,3..8 circle with a break row(_break_=_page_) in every 8 rows. However in SAS on my computer, the out= dataset shows: Obs SALES DATE YEAR MONTH DAY haha _BREAK_ .......... 41 $606 90Q1 1990 1 1 41 42 $674 90Q2 1990 4 1 42 43 $705 90Q3 1990 7 1 43 44 $749 90Q4 1990 10 1 44 45 $703 91Q1 1991 1 1 45 46 $709 91Q2 1991 4 1 46 47 $728 91Q3 1991 7 1 47 48 $807 91Q4 1991 10 1 48 49 $692 92Q1 1992 1 1 49 50 $797 92Q2 1992 4 1 50 51 $826 92Q3 1992 7 1 51 52 $889 92Q4 1992 10 1 52 53 $758 93Q1 1993 1 1 53 54 $909 93Q2 1993 4 1 54 55 $920 93Q3 1993 7 1 55 56 $991 93Q4 1993 10 1 56 57 $876 94Q1 1994 1 1 57 58 $998 94Q2 1994 4 1 58 59 $998 94Q2 1994 4 1 _PAGE_ only one _break_=_page_ in the end. haha does not match what it is in the report. It seems like the first compute after _page_ block is completely ingored in running out= dataset, and I have no idea why this is happpening. I hope my broken English can make you understand my confusion here. And I will also try to consult with Tech support. Thanks again for your reply.
... View more