I believe that the problem is, at least for Tabulate, that the headers are technically always character - so in creating the resulting xml, they're sent as 'string'.  You might be able to fix this by editing the tagset itself, though I'm not sure where I'd suggest beginning there.  I don't think tagsets.excelxp will receive too much more development now that ODS EXCEL is largely out, so I wouldn't expect it to change - but who knows. ... View more

I'm not convinced it's working, in any event; looking at the raw XML, it's still a string - excel's just being smart about it.  I am probably not using the style properly, I am a PROC REPORT guy not a PROC TABULATE guy when it comes to these sorts of things. ... View more

Hmm, or maybe not.  Does this work the way you want it to? edit: I think this is sort-of-working, but the format still changes when you select the cell and hit 'enter' - but at least it's closer.  Not sure if this is the 'does not work' I remember earlier or if this is an improvement. IE, if you go to cell D1, select, hit enter, it goes from "02SEP86" to "2-Sep-86".  But, at least it's a real date value you can do math on and such. proc template; define style test; parent=styles.normal; class date/   tagattr="m/d/yyyy"; end; quit; ods listing; options missing = 0; ods tagsets.excelxp file= "C:\temp\test.xml"   style=test; proc tabulate data=sashelp.stocks missing; class stock date/style=data_date9; var high low; tables (high low),date=''*{style=date} /row = float; run; ODS TAGSETS.EXCELXP CLOSE; ... View more

I don't think so.  See my first response.  I don't have the code I used to reverify this when this was originally posted, but from what I recall, Excel won't display the date properly until the cell is selected and validated (hit enter).  This was the case for me a year or two ago when I first hit this issue, and I verified it back when this question was originally posted. ... View more

I'm fairly sure there is no good answer, short of modifying the template or some other hack - it's an oddity in how excel works relating to the open xml format.  ODS EXCEL may fix this, once I can get my hands on a copy of 9.4 TS1M1 I will confirm (probably not for a few weeks), but that obviously doesn't help a 9.3 user. ... View more

&DT would need to resolve to "16Dec2013"d, not 16Dec2013 with no quotes or d ("xxx"d is a "date constant").  It's possible in the past the macro variable was defined as the numeric equivalent of the date, but now you're trying to use the text (which is fine). data _null_; suffix = put("&DT"d,yymmddn8.); call symput('DATE', suffix); run; Otherwise, you could consider trying something like %let date=%sysfunc(today(),YYMMDDN8.); if it's intended to be today, for example; other similar solutions exist for specifying particular dates, depending on how your particular date is arrived at. ... View more

I would look into DDE, if you can (google "SAS Excel DDE").  That allows SAS to tell Excel to do things, including run macros. ... View more

I first off would suggest not having a dataset with months as variables.  Store it in normalized form with months as rows.  Then you can easily select the last 6 months' worth of data using a WHERE or subsetting IF statement, and if needed transpose back. If you must keep it in horizontal format, you could select the variable names from dictionary.columns (or proc contents output) into a macro variable.  Substitute proper names for LIBNAME and MEMNAME below (MEMNAME = dataset name). proc sql; select name into :collist separated by ' ' from dictionary.columns where libname='WORK' and memname='HAVE' and input(name,MONYY.) ge intnx('Month','01NOV2013'd,-5); *you can easily adjust this to be programmatic, not sure what your conditions are for determining 'last month' - use INTNX to pick last 6 months once you have that defined [macro variable, last actual calendar month, etc.]; quit; data want; set have; keep &colllist.; *and any other needed variables.; run; ... View more

I ran into the same problem, actually, a while back.  If I recall correctly, I think I fixed it by installing in compatibility mode for windows 7. And agreed, SAS uses a strange and poorly managed installation method.  No idea why they don't just use MSIs or Installshield or whatnot.  9.4 and 9.3 are much better than 9.1 was (... runs screaming ...), but still not the best installation experience. ... View more

I'm not sure what the problem is, there.  That's what you'd expect from that (perhaps the doc can be more clear, certainly).  With continuous, it's counting boundaries, true, but how else would you calculate the number of days between something?  You'd have to do some sort of boundary.  The point with continuous (which is horribly named; INTNX 'same'/'s' makes far more sense) is that it counts truly the number of [whatevers] since [exact date or datetime], not the number of crossings of an unrelated boundary (ie, midnight, or Jan 1, or whatnot.) ... View more

Note that if you use 'C' argument at the end, it will count continuously rather than across boundaries.  (This is what PGStats was obliquely referring to.) ... View more