Thanks. I didn't know "constant('big')" of course, hence the algorithm. ... View more

An example use case: Data being read has a date field in mm/dd/yyyy format but missing values where the data was not recorded are represented by 99/99/9999. Since you know that values will be invalid dates and actually represent missing data when the values are read you will be okay with SAS variables with missing values but you may not want to wade through a bunch of junk in the log for a know data issue. So a data step to read the example data uses the ?? modifier as follows:   data example; input x ??mmddyy10.; format x mmddyy10.; datalines; 01/01/2019 99/99/9999 ; run; remove the ?? and replace with either a single ? or nothing to see the different behaviors.   Reading a field that should be numeric but sometimes has NA or other not entered or not available code would be another similar use.   ... View more

If you have strings then locate the decimal point and add the appropriate number of leading spaces. data have; input str :$10. @@; cards; 121 6.02 42 42 10.08 67 31 7.08 88 122 3.24 3.24 10.44 2.88 3.24 3.24 78 89 92 78 85 9.72 11 104 92 64 7.2 ; data want ; set have ; length varout $10.; format varout $char10.; loc = index(str,'.'); if loc=0 then loc=length(str)+1; varout=str ; if loc < 8 then varout=cat(repeat(' ',7-loc),varout); run; proc print; run; Obs str varout loc 1 121 121 4 2 6.02 6.02 2 3 42 42 3 4 42 42 3 5 10.08 10.08 3 6 67 67 3 7 31 31 3 8 7.08 7.08 2 9 88 88 3 10 122 122 4 11 3.24 3.24 2 12 3.24 3.24 2 13 10.44 10.44 3 14 2.88 2.88 2 15 3.24 3.24 2 16 3.24 3.24 2 17 78 78 3 18 89 89 3 19 92 92 3 20 78 78 3 21 85 85 3 22 9.72 9.72 2 23 11 11 3 24 104 104 4 25 92 92 3 26 64 64 3 27 7.2 7.2 2 ... View more

It looks like ADT is the variable that can order the observations by date.  (If you prefer to use VISITNUM to do this, that's OK.)   Here's a way:   proc sort data=have; by usubjid paramcd adt; run;   data want; update have have (keep=usubjid paramcd ablfl); by usubjid paramcd; output; run;   It's untested code, but should be fine. ... View more

First, please post test data - i.e. a small sample of test data which illustrates your problem, post this test data in the form of a datastep: https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-data-AKA-generate/ta-p/258712   Do not post real data or Excel files.   Next up, show what you want out from this test data.   At a glance at what you have posted there, you don't want a "tranpose" at all.  You basically want to merge some data back onto the main data, something like: data want; set have have (where=(lbtestcd="ALT" and visit="SCREENING") rename=(lbstresn=alt ...) keep=usubjid alt) have (where=(lbtestcd="AST" and visit="SCREENING") rename=(lbstresn=ast ...) keep=usubjid ast); by usubjid; run; Note this just merges on baseline to each record, you would need to be far more specific about your issue for any more accurate examples and provide usable test data. ... View more

This is a task for a scheduling system, not SAS.   In SAS, use the exist() function to check for a dataset, and the sleep() function to have a program lie waiting for a given time. ... View more

I would look into the sleep() function. It delays execution by the time specified.    Here is SAS's documentation on the function: http://support.sas.com/documentation/cdl/en/lefunctionsref/69762/HTML/default/viewer.htm#p0a6vn2td7bjr2n1viy8y4lgvq61.htm   I quickly wrote up a crude macro that could be used for this. Perhaps there's a more creative method, but this works.     /******************************************************************* Macro to delay processing: Number: Amount of time to wait Units: Unit of time to wait ("seconds", "minutes", "hours", "days") ********************************************************************/ %macro waiting(number, units); data wait; if &units. = "seconds" then unit = 1; *Base unit in sleep function; else if &units.="minutes" then unit=60; *60 seconds in a minute; else if &units.="hours" then unit=3600; *3,600 seconds in an hours (60 seconds per minute * 60 minutes per hour); else if &units.="days" then unit=86400; *86,400 seconds in a day (60 seconds per minute * 60 minutes per hour * 24 hours per day); wait_time = sleep(&number., unit); run; %mend; %waiting(number=10, units="minutes");   ... View more

When you issue a lock() (or whatever the system call is in any given operating system), the system sets the lock for the calling process. As long as you try to modify the file from the same process (read: same workspace server instance), the lock will not be effective; it just blocks other processes from accessing the file. If you want to see the effect of a lock on another process, just add a sleep call in your code: lock AccesChk.test01 ; data _null_; x1 = sleep(600,1); run; lock AccesChk.test01 clear ; Now you can test if the file is prevented from access in the other session. ... View more

You didn't post the log so hard to see if there was an issue with the way you submitted the code. Perhaps the lines were too long and became truncated?   But most likely an earlier error has put SAS into a mode where it is not running anyghing. I would suggest looking higher in the log for the error that actually caused SAS to set OBS=0. ... View more

I normally restrict my search to support.sas.com   The following Google term should give you quite a good result: site:support.sas.com 9.4 call symput ... View more

This sample is similar to what you want to do.   http://support.sas.com/kb/51/626.html   I think you will need to figure how to make each "page" of your TEXT file a separate document so you will have proper pagination. ... View more