proc compare check on observation level meaning check osb 1 from clm is equal to obs 1 from gsl. I unterstand your task is to catch observations which have no match in the opposite table. So my solution is: proc sort data = clm; by counterparty; run; proc sort data = gsl; by splr_name; run; data clm_only (rename =(joinvar = counterparty ) drop = relid phn) gsl_only (rename =(joinvar = splr_name ) drop = id name); merge clm (in =inclm rename =(counterparty = joinvar)) gsl (in =ingsl rename =(splr_name = joinvar)); by joinvar; if ingsl eq 0 then output clm_only; if inclm eq 0 then output gsl_only; run; ... View more

You may use a datastep or proc datasets / proc copy instead: data LIB_SQL.WXGC21K; set backup.WXGC21K ; run; ... View more

The schema is needed in LIBNAME statement. For pass through the tablespace follows the FROM statement and then a period + table name:   proc sql; connect to SQLSVR as mydb (Datasrc="EDWHDSST" AUTHDOMAIN="SQLSRVAuth" ); execute (SELECT * FROM dbo.TABSQL) by sqlsvr; disconnect from sqlsvr; ... View more

@Ksharp showed the answer of your question already : ":" works as a wildcard so all tables starting with "table_" will be used to create the result. I made your example runable so you may test it: data table_1; length value $10.; input value; datalines; aaa bbb ; run; data table_2; length value $10.; input value; datalines; aa bb ; run; data table_3; length value $10.; input value; datalines; a b c ; run; data final ; set table_: ; run; ... View more

I do not know much about the environment you are working in. I assume that proc append just take a "few" observations and gives it to SQLServer who appends them. After doing so maybe the SQLServer starts re-creating  an index or so, waits for completion and after that accepts the next package.   I would try to get the data to be appended in the SQLServers database and let the SQLServer doing the append.  Something like this  :  proc upload data = backup.WXGC21K out = LIB_SQL.WXGC21K;run; proc sql; connect to ODBC (dsn="MySQLServerConnectString" schema=dbo); execute (INSERT INTO TAB1 SELECT * FROM dbo.WXGC21K) by ODBC; disconnect from ODBC; quit; proc sql; connect to ODBC (dsn="MySQLServerConnectString" schema=dbo); execute (Drop Table WXGC21K) by ODBC; disconnect from ODBC; quit;; or if you have licenced SAS/Interface for Access to SQLServer you may try this:   proc upload data = backup.WXGC21K out = LIB_SQL.WXGC21K;run; proc sql; connect to sqlsvr (user="Domain\UserID" password="MyPassword" datasrc="MySQLServerConnectString" schema=dbo);; execute (INSERT INTO TAB1 SELECT * FROM WXGC21K) by sqlsvr; disconnect from sqlsvr; quit; proc sql; connect to sqlsvr (user="Domain\UserID" password="MyPassword" datasrc="MySQLServerConnectString" schema=dbo); execute (Drop Table WXGC21K) by sqlsvr; disconnect from sqlsvr; quit;; ... View more

proc import out=test datafile='C:\data\test.xlsx' dbms=xlsx replace; getnames=yes; run; data test (drop = test_name_import); set test (rename =(Test_name = test_name_import)); format test_name best12.; test_name = input(test_name_import, comma32.); run; ... View more

Understand. But then the solution provided by @SASKiwi in combination with the check provided by @Quentin should work. Maybe it's necessary to quote the full path to sas.exe (mind the double quotes & take your path not my 32bit one🤐 & the temp folder should exist already) in the shortcut - like this "C:\Program Files\SASHome\x86\SASFoundation\9.4\sas.exe" -work "E:\temp" -autoexec "C:\Users\...\autoexec.sas" -config "C:\Users\...\sasv9en.cfg" ... View more

From my point of view the problem is free disk space (110 GB) is lower than dataset size (115GB). When SAS sorts it creates a new file with the extension ".lck" to write the observations in sorted order. When it's finished the unsorted file is deleted and the new file is renamed (cut off the trailing ".lck"). This means for a very short period of time you need the dateset size twice (2 x 115 GB). That's why you run in the "disk full condition".  My solution is create a library on E: , move (not copy) the unsorted dataset in there, sort it there and move it back or use the out= option in proc sort to write the sorted dataset on C: directly     ... View more

data Treatment1; format Subject $4. Visit $10. Visit_date date10. NextCycleManuallyAdded $3.; input Subject $8. Visit $12. Visit_date MMDDYY10. NextCycleManuallyAdded $5.; cards; A001 Screening 10/27/2021 Yes A001 cycle1 12/1/2021 Yes A001 cycle2 1/17/2022 Yes A001 cycle3 2/28/2022 Yes A001 cycle4 4/19/2022 Yes A001 cycle5 5/30/2022 Yes A001 cycle6 7/11/2022 Yes A001 cycle7 8/22/2022 Yes A001 cycle8 10/3/2022 Yes A001 cycle9 11/17/2022 Yes A001 cycle10 1/2/2023 Yes A001 cycle11 2/15/2023 Yes A001 cycle12 3/27/2023 Yes A002 Screening 11/22/2022 Yes A002 cycle1 1/14/2023 Yes A002 cycle2 2/28/2023 Yes A003 Screening 9/30/2022 Yes A003 cycle1 11/2/2022 Yes A003 cycle2 11/23/2022 Yes A003 cycle3 1/2/2023 No A003 EOT 1/9/2023 A004 Screening 9/1/2022 Yes A005 cycle1 9/23/2022 Yes A005 cycle2 10/31/2022 No A005 EOT 11/2/2022 ; run; proc sort data = Treatment; by subject descending visit_date; run; data Treatment2; set Treatment; format NextCyle _TempNextCyle $3.; retain _TempNextCyle; by Subject; select; when(first.Subject and Visit eq 'EOT' ) do; NextCyle = ''; _TempNextCyle = 'No';end; when(first.Subject) do; NextCyle = 'Yes'; _TempNextCyle = 'Yes';end; otherwise do; NextCyle = _TempNextCyle; _TempNextCyle = 'Yes';end; end; /* select; when(NextCyle eq NextCycleManuallyAdded) CorrectResult = 1; otherwise CorrectResult = 0; end; */ drop _: ; run; proc sort data = Treatment2; by subject visit_date; run; ... View more