If a blank to start/end the string may not work try a null charcter. In Windows at least, hold down the Alt key and press 255 on the numeric key pad. It looks like a blank when printed but doesn't normally get stripped in other places, so may not in this case as well. ... View more

> Hey KSharp, > Is there a way to use this to create a numeric > variable instead of a character one? or do i just > need to convert it back to a numeric after i'm done? > > Thanks, > cr Not the best programming practice but if you leave off the length declaration the target variable will default to numeric and you'll get a numeric result and warnings in the log about conversion from string to numeric. ... View more

First make sure that the two sets are sorted the same. You may also want to set a FUZZ value of something like 0.000001. I've had lots of cases comparing aggregated data back to an original data set where the differences were in the 1E-13 range and PROC COMPARE will set tell you they are all different. So either FUZZ or round values may help. ... View more

How are you setting the option? ... View more

> You can also bypass this problem by running the > program in batch mode: from a command line type SAS > -SYSIN "mysasprog.sas" Which should also generate a log file you can review at need. ... View more

Depending on what you mean per Cynthia's questions, I have had occasion to want both a code and text at the same time. My approach has been, in the steps that create the input data set, to create a new text variable that is the concatenation of the code and the text. Pick the order to concatenate depending on how you want the sort order in the tables. ... View more

You may have a preference (Tools -> Options -> Preferences in Windows) set for "Save settings on exit". ... View more

If I were more interested in the results than the particular Proc I'd try without any other manipulation of the Test data set. proc tabulate data=test; class birthyear married divorced6; var work; table birthyear all='Total', ( divorced6 married) *work='' *(n='Count' mean*f=f8.4); run; ... View more

Single input file or multiple input files with varying numbers of comment lines? ... View more

Or modified approach using CLASS statement instead of BY. This doesn't require input data to be sorted. proc summary data=orig_srt nway; class id; var activity; output out=orig_sum(drop=_type_ _freq_) sum=; run; ... View more

> Hello RMan, > > There is a macro function %TRIM. May be your problems > result from this confusion? > > Sincerely, > SPR That wouldn't be it. Most of the system macro "functions" hang out, in my install, at C:\Program Files\SAS\SASFoundation\9.2\core\sasmacro or similar. ... View more

> Hello, > > I have two related problems. > > 1) I have a data set with 100 observation I only need > the first 15 of them. on the set statement ( or just about anywhere you have a DATA= use the OBS= option Data newdata; set olddata (obs=15); run; Also look at FIRSTOBS. > > 2) next only need 5 variables of the 14 variables > > I used KEEP statement but my data came out with the > wrong order. These are binary data and I need to keep > the order the way it was in the original data. Which order? Sort, as in row order, or columns in a dataset. If the latter I wouldn't worry about as other applications that may want the data would either link to the variable name or when you export the data then specify the column order. ... View more

MAUTOSOURCE option set? ... View more

Here are a couple of other things to add to the KEYS menu: If you use a lot of PROC PRINT just to look at the last data created try this: vt _last_ mode=edit; To clear out the results, output and log window and return to the enhanced editor: odsresults;select all;clear;log;clear;wpgm; ... View more