SAS Support Communities

ballardw - I apologize if my response seems terse. SAS 9.4 is already installed on the client machines alongside 9.3.   I guess I'm perturbed because, in my experience, SAS has never forced us to switch to a new version while also expiring the previous version. Perhaps my memory is bad.   In any case, this has resulted in weeks of work that was not on my workplan - in a year that was already overflowing with work demands, To add insult to injury, the final step (uninstalling SAS 9.3) results in 9.4 failure and then more hours of unplanned work.   SAS needs to fix their buggy uninstaller, rather than forcing their customers to deal with it.  ... View more

DJDaniel - in future, would this procedure solve the problem of an uninstall of a previous version of SAS (e.g. 9.3) when a later version is present (e.g 9.4)? ... View more

I appreciate the feedback, but you haven't answered my question. Uninstalling 9.3 first means I have to completely configure 9.4 from scratch. ... View more

Thanks for the reply, Juan. As for AMO, we are using the latest 9.4 standalone installer and it detects and installs the correct version based on Office bitness.   My "problem" is that I don't understand what SAS-Office "interoperability" means with respect to Base SAS/EG. As I replied to Patrick, I am able to import/export to/from SAS/Excel regardless of bit match - without any special code, which seems to contradict Chris' blog posts. ... View more

Thanks, Patrick. I'm aware of Chris' blog and have read all of his articles (including the one you linked). The "problem" is that I am able to get results (import/export between SAS and Excel to and from 32-bit and 64-bit) without using special code/methods as Chris describes.   Still investigating, so stay tuned! ... View more

Here's a sample that I've adapted from SAS support examples online. It uses PROC TABULATE with the PRELOADFMT and PRINTMISS options and OUT= to create a table with all possible combinations of class variables. I'm sure you'll get the idea from this code sample.   (In your specific case, the FORMAT you should create would be for fiscal year. Obviously, fiscal year isn't in this example. It's just to illustrate how you can solve the problem of having missing classification levels. There are other approaches that could also be used . . . .)   data energy;    length State $2;    input Region Division state $ Type Expenditures;    datalines; 1 1 ME 1 708 1 1 ME 2 379 1 1 NH 1 597 1 1 NH 2 301 1 1 VT 1 353 1 1 VT 2 188 1 1 MA 1 3264 1 1 MA 2 2498 1 1 RI 1 531 1 1 RI 2 358 1 1 CT 1 2024 1 1 CT 2 1405 1 2 NY 1 8786 1 2 NY 2 7825 1 2 NJ 1 4115 1 2 NJ 2 3558 1 2 PA 1 6478 1 2 PA 2 3695 4 3 MT 1 322 4 3 MT 2 232 4 3 ID 1 392 4 3 ID 2 298 4 3 WY 1 194 4 3 WY 2 184 4 3 CO 1 1215 4 3 CO 2 1173 4 3 NM 1 545 4 3 NM 2 578 4 3 AZ 1 1694 4 3 AZ 2 1448 4 3 UT 1 621 4 3 UT 2 438 4 3 NV 1 493 4 3 NV 2 378 4 4 WA 1 1680 4 4 WA 2 1122 4 4 OR 1 1014 4 4 OR 2 756 4 4 CA 1 10643 4 4 CA 2 10114 4 4 AK 1 349 4 4 AK 2 329 4 4 HI 1 273 4 4 HI 2 298 ; ;;;;   proc format;    value regfmt 1='Northeast'                 2='South'                 3='Midwest'                 4='West';    value divfmt 1='New England'                 2='Middle Atlantic'                 3='Mountain'                 4='Pacific';    value usetype 1='Residential Customers'                  2='Business Customers'; run;   proc tabulate data=energy format=dollar12. out=outtab;    class region division type / preloadfmt;    var expenditures;    table region*division,          type*expenditures / rts=25 printmiss;    format region regfmt. division divfmt. type usetype.;    title 'Energy Expenditures for Each Region';    title2 '(millions of dollars)'; run;   . . . . then proc transpose outtab . . . .  ... View more

A relatively straightforward solution is to ensure that the final dataset you are transposing has all possible levels of the variable, by merging them in prior to the transpose. ... View more

With regard to questions 1 & 2, have you looked at ODS Excel XP options? http://support.sas.com/rnd/base/ods/odsmarkup/TipSheet_ods_xl_xp.pdf ... View more

If you have a large number of values for "Area" and "Chart_Field_1", you can use a macro to loop through the possible values and output a corresponding Excel workbook for each value of "Area" containing separate spreadsheets for each value of "Chart_Field_1". To give you some ideas, look at this post: https://communities.sas.com/t5/SAS-Procedures/Split-SAS-Output-into-multiple-Sheets/td-p/18996 If the number of variables is small, you could hard-code them. ... View more