BookmarkSubscribeRSS Feed
HimaAalamuri
Fluorite | Level 6

Hi all,

i want to compare output window results, is there any procedure is there to compare output windows?

Best Regards,

Hima

9 REPLIES 9
LinusH
Tourmaline | Level 20

Please describe in detail of what you are doing, and why (the underlying requirement).

Data never sleeps
HimaAalamuri
Fluorite | Level 6

Hi Linus,

i am developing one table (say x) and coming that table with production result by using proc compare, and i have to generate one more listing which contains output window results for x. so i have to compare production listing with my QC listing! my question is how i compare by manually or is it any procedure is der to compare output windows?

thanks in advance!!

Hima

RW9
Diamond | Level 26 RW9
Diamond | Level 26

There are several ways, but not via output window.  You can route your data to datasets as Xia Keshan has stated, and this then gives you proc compare options.  That in itself takes a lot of the QC out from the output file.  AS for your actual output, what do you need to check exactly as styling and such like can make differences.  If you just want basic content, then use a proc print to text file, then you can read that in.  For more advanced options then you can go out to RTF or PDF and use that software's compare functions.  You could also read the rtf back in and parse it into its composite parts, I have done this and its not particularly difficult, the main issue is parsing all the tags and such like.  You could also set your output window up to display HTML and then save that to file, then create a HTML parse program to read it back in.

Me, I would say that effort vs confidence ratio would suggest that proc compare the data as a "one step away" dataset, i.e. make your data look almost identical to the output, and then having a small output routine would be most cost efficient.

So if you have a table which looks like:

Category                                   Trt1          Trt2         

                                                  (n=xx)      (n=xx)

---------------------------------------------------------------

ABC                                           xx (yy%)  xx (yy%)

Then make a dataset which looks like that:

CAT (Category)     TRT1 (Trt1|(n=xx))         TRT2 (Trt2|(n=xx))

ABC                       xx (yy%)                         xx (yy%)

Store that, get the QC'er to do the same, and proc compare the two.  Then a manual review of the output to ensure that the right dataset was used, and titles/footnote are correct.

HimaAalamuri
Fluorite | Level 6

thank you RW9,

But my problem is i am using proc mixed in generating table visits wise so i got around 40 pages in output window i have to compare my output window to production programer output window which is in PDF file.

Thanks in advance!

RW9
Diamond | Level 26 RW9
Diamond | Level 26

Well, then your stuck really.  As I have stated, direct your output not to the output window - which to be fair is about as useful as a hole in a bucket.  Just put an ods statement before and after your mixed and send to a pdf file.  You can then use PDF compare to check the two.  However you need to bear in mind that PDF compare will check everything, so if your not using the same style or layout then it will not match.  Hence I recommend that you are proc comparing datasets which is the best way to get accurate comparison of the data.  If you want to keep your mixed output then use ods select for the particular tables which mixed creates behind the scenes:

http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_mixed_sect02...

Each of these holds some of the statistics produced and can be kept.

HimaAalamuri
Fluorite | Level 6

Thank you RW9,

please see the below path and let me know ur valuable replay!!

https://communities.sas.com/thread/65129

RW9
Diamond | Level 26 RW9
Diamond | Level 26

Well, not too familiar with GLM, but here is a SAS article which shows sending output from those types of procedures to html, its pretty much the same thing for PDF:

http://support.sas.com/documentation/cdl/en/odsug/61723/HTML/default/viewer.htm#a002233364.htm

Also note, that to find what output blocks are created in your procedure, use the ods trace on; syntax, this will display in the output window what objects are created, you can then re-run the program select those objects to send via ods to a file.

HimaAalamuri
Fluorite | Level 6

Ok No problem thank you Smiley Happy

Ksharp
Super User

input those output into sas datasets , and using proc compare .

You can also compare them by the third software .

sas-innovate-2024.png

Join us for SAS Innovate April 16-19 at the Aria in Las Vegas. Bring the team and save big with our group pricing for a limited time only.

Pre-conference courses and tutorials are filling up fast and are always a sellout. Register today to reserve your seat.

 

Register now!

What is Bayesian Analysis?

Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.

Find more tutorials on the SAS Users YouTube channel.

Click image to register for webinarClick image to register for webinar

Classroom Training Available!

Select SAS Training centers are offering in-person courses. View upcoming courses for:

View all other training opportunities.

Discussion stats
  • 9 replies
  • 1156 views
  • 1 like
  • 4 in conversation