I’d ask what’s causing this error in first place. Ideally you’ve got some error handling in your job so it doesn’t go into syntax check mode.  If an error situation doesn’t cause this table to not even get populated then you could of course also set options obs=max for the portion of your code where you create and send the email using some code along the line of below (not tested, but syntax will be at least close to working). %let sv_obs=%sysfunc(getoption(obs,keyword)); Options obs=max; /* here your code for the email */ Options &sv_obs; ... View more

From the looks of it and to avoid the root cause for warnings as explained by @ChrisNZ I believe one of below coding options should return what you're after. proc tabulate data=s_18 format=nlnum32. out=s_18_CF; class original_currency yearno company_name company_bk company_type_code period_name typeno; var epifp; table Company_Name *company_bk ,yearno='Year' *company_type_code='Institutgroup' *original_currency='Curr' *period_name='Period' all='Total' ,all='Total' *epifp ; run; proc tabulate data=s_18 format=nlnum32. out=s_18_CF; by Company_Name company_bk; class original_currency yearno company_type_code period_name typeno; var epifp; table yearno='Year' *company_type_code='Institutgroup' *original_currency='Curr' *period_name='Period' all='Total' ,all='Total' *epifp ; run;     ... View more

@Cykelops As others already stated variable grouping doesn't exist in your source data. From the looks of it you just want to create a frequency table for every single variable in your data source. You can use keyword _all_ to address all existing variables. proc freq data = CARI_ICDW_PB ; tables _all_; run; ... View more

Besides of using Proc Contents to ensure that the variable actually exists there is also a semicolon missing in the code you shared after variable grouping. ... View more

You could add some error handling for such cases. Something like: if weekday(date2) in (1,7) then do; /* here some error handling code */ end; You could for example ensure that your time_stores variables become missing if date2 falls on a weekend and though is invalid. Or you could  "fix" the data and shift it on to the next Monday. data want; set have; format date1 date2 shift_date1 shift_date2 weekdatx.; shift_date1=date1; shift_date2=date2; if weekday(shift_date2) in (1,7) then do; /* Option1: set shift_date2 to missing so derived variables become missing */ /* shift_date2=.; */ /* Option2: "fix" data by shifting date2 to Monday */ shift_date2=intnx('weekday17w',shift_date2,1); end; /* if shift_date1 a Saturday or Sunday shift it to Monday */ if weekday(shift_date1) in (1,7) then shift_date1=intnx('weekday17w',shift_date1,1); derived_timetores_1=shift_date2-shift_date1; /* derived_timetores_1=intck('day',shift_date1,shift_date2); */ derived_timetores_2=intck('weekday17w',shift_date1,shift_date2); run; ... View more

I don't really understand how your process/flow looks like but based on your description it sounds like a timing issue. Make sure that your error check/send email step is outside of the loop (especially when running in parallel).  ... View more

@sasgorilla wrote: Thanks, Patrick! This appears to have worked except there is one particular observation that is (-) for both values, (-2) for want1 and (-1) for want2. This is the only negative value out of a couple thousand observations.    For this observation, Date1=04/09/2022 and Date 2=04/09/2022;    Any ideas on how to troubleshoot to figure out what's causing this?  That's because your Date2 is on a Sunday which is obviously not a working day and though based on your initial definition shouldn't be possible.  If you have to cater for such cases then you need to provide amended sample data that covers such cases, shows the desired result and explains the logic. /* timetores_1: don't count only 1st weekend if date1 on 1st weekend */ /* timetores_1: never count weekend */ data have; input date1 :yymmdd10. date2 :yymmdd10. timetores_1 timetores_2; format date1 yymmdd10. date2 yymmdd10.; datalines; 2024-11-02 2024-11-04 0 0 2024-11-04 2024-11-04 0 0 2024-11-05 2024-11-07 2 2 2024-11-07 2024-11-11 4 2 2024-11-09 2024-11-18 7 5 2022-09-04 2022-09-04 0 0 ; run; data want; set have; format date1 date2 shift_date1 weekdatx.; shift_date1=date1; /* if shift_date1 a Saturday or Sunday shift it to Monday */ if weekday(shift_date1) in (1,7) then shift_date1=intnx('weekday17w',shift_date1,1); derived_timetores_1=date2-shift_date1; /* derived_timetores_1=intck('day',shift_date1,date2); */ derived_timetores_2=intck('weekday17w',shift_date1,date2); run; proc print data=want; run;     ... View more

@sasgenie wrote: The trick is to prevent SAS from showing the html output in its own output window, and use a browser for output instead. Honestly based on the "solution," it looks like an issue easy to fix and I don't understand why SAS cannot do it.   Follow these steps: Tools -> Options -> Preferences -> Results table -> View results using Chrome One reason might be that that the "PC SAS" Window Manager environment is obsolete since many years and that there isn't any further development on it. Start using SAS EG or even better SAS Studio. You could even use Visual Studio Code with the SAS extension or Jupiter Notebook if that's your preference. ... View more

It's always really helpful if sample data gets provided. Assuming I did create correct sample data below should work. data have; input event_dt:date9. resolve_dt:date9. days2resolve_want; format event_dt resolve_dt date9.; datalines; 15nov2024 18nov2024 1 16nov2024 18nov2024 0 17nov2024 18nov2024 0 18nov2024 18nov2024 0 16nov2024 19nov2024 1 17nov2024 19nov2024 1 18nov2024 19nov2024 1 18nov2024 20nov2024 2 15nov2024 26nov2024 7 16nov2024 26nov2024 6 17nov2024 26nov2024 6 18nov2024 26nov2024 6 19nov2024 26nov2024 5 ; data want; set have; if weekday(event_dt) in (1,7) then days2resolve_derived=intck('weekday17w',intnx('weekday17w',event_dt,1),resolve_dt); else days2resolve_derived=intck('weekday17w',event_dt,resolve_dt); run; proc print data=want; run;     ... View more

Not really sure what your problem is but look into SAS system option Locale and if using SAS Studio also into your browser settings. Please share a bit more detail for us to be certain that this is a genuine SAS question and not just spam.   ... View more

Below will write the first occurence of an ID to table WANT. The total of 5 is then simply the number of rows in table want. data want; if _n_=1 then do; dcl hash h1(); h1.defineKey('patid'); h1.defineDone(); end; set test2016 - test2020; if h1.check() ne 0 then do; output; h1.add(); end; run; proc print data=want; run;     ... View more

The issue is:   You would have to use something like powershell in your pipe command. The default cmd.exe simply doesn't support UNC paths. This is a Microsoft thing and nothing you can change. But... There is really no need to use an external command which requires XCMD to list files and especially SAS tables. Just use an approach along the line of what @Ksharp already proposed.   If you want to use your current code as-is then you need to map this UNC path (=give it a drive letter) on the machine where SAS executes. ... View more