I agree that 08022024 can be interpreted as August 2 or February 8, but no way for 20240208. 🙂   Bart ... View more

Thank you very much ! ... View more

Please share data as text. (why go to the trouble to type it into Excel and then upload a file when you can just post the SAS code???) data have; input NO $ MT_EA jour_mois $ IS_PROTOCOLE $ LP_ETAT_DOSS $ CD_RGA $; cards; A020 31413.16204 31/1 1112 VALID FR10 A020 31289.79011 28/2 1112 VALID FR10 A020 31396.92356 31/3 1112 VALID FR10 A020 31406.90669 30/4 1112 VALID FR10 A020 31413.5621 31/5 1112 VALID FR10 A020 31304.2403 30/6 1112 VALID FR10 A020 31390.58766 31/7 1112 VALID FR10 A020 31480.25608 31/8 1112 VALID FR10 A020 31516.72788 30/9 1112 VALID FR10 A020 32580.65359 31/10 1112 VALID FR10 A020 32620.42652 30/11 1112 VALID FR10 A020 33130.84571 31/12 1112 VALID FR10 A020 5000 31/1 1112 VALID FR20 A020 6000 28/2 1112 VALID FR20 A020 6000 31/3 1112 VALID FR20 A020 6000 30/4 1112 VALID FR20 A020 6000 31/5 1112 VALID FR20 A020 6000 30/6 1112 VALID FR20 A020 6000 31/7 1112 VALID FR20 A020 6000 31/8 1112 VALID FR20 A020 6000 30/9 1112 VALID FR20 A020 6000 31/10 1112 VALID FR20 A020 6000 30/11 1112 VALID FR20 A020 6000 31/12 1112 VALID FR20 ; From your output it looks like you just want to transpose ONE variable.  But the PROC SUMMARY code you posted was trying to transpose TWO variables.   If you only want to transpose ONE variable then PROC TRANSPOSE will work fine. proc transpose data=have out=want(drop=_name_) delim=_; by no; id cd_rga jour_mois ; var mt_ea ; run; Result: PS Make sure you have not accidentally set the VALIDVARNAME option to ANY or else those awful slashed in the JOUR_MOIS variable will get inserted into the generated variable names. options validvarname=v7; ... View more

Thank you, Tom ! ... View more

Thank you, Tom ... View more

It's called Maxim#4   proc sql; create table t12_sql as ( select t1.* from T1 left join T3 on t1.A=t3.A ) union ALL select * from T2 ; quit; ... View more

data results; set pdf_corr; length src dst dst2 $ 8 ; *create path to files; source_file =catt("\\xxxxx\PDF_20230427\", NOM_PDF1); dest_file = catt("\\xxxxx\PDF_20230427_V2\", NOM_PDF1); *create filerefs for copy function; dest_file2 = dest_file = catt("\\xxxxx\PDF_20230427_V2\", NOM_PDF2); rc1=filename(src,source_file,,'recfm=n'); rc2=filename(dst,dest_file,,'recfm=n'); rc2_1 = filename(dst2, dest_file2, , 'recfm=n'); rc3_1=fcopy(src,dst); *deassign filerefs; rc3_2=fcopy(src,dst2); *deassign filerefs; rc4=filename(src); rc5=filename(dst); rc6 = filename(dst2); run; Assuming you have another variable with the name, then the above is a brief example. If every file is copied multiple times you can generate those names and create a loop to loop over the names rather than repeat the code but it depends on exactly what you're doing. ... View more

Thank you Koen! ... View more

Not sure I understand.  How can you add observations if the datasets have different variables?  That are you putting into the new observations in that case?   Provide a clearer example.   Perhaps you just need to make the subset of the second dataset first before trying to append/merge it with the first dataset? ... View more

Sorry, It is actually day (friday) of last week before the date. ... View more

Then you have to use INTNX or INTCK with a DTDAY interval. ... View more

Suggestion: Pick dates that you know the expected result for "age". Put those into the function and see the result.   If you don't know how to put a specific data the structure is 'ddMONyyyyy'd where MON is the 3-letter abbreviation of the month.   So yrdif ('02JAN2015'd, '15JUL2023'd) is what yrdif returns for a date of birth of 02JAN2015 with age evaluated on 15Jul2023. ... View more

Are you sure the issue is the COALESCE() function call?  It seems to work fine for me 570 proc sql; 571 create table test as select a.*,coalesce(a.height,a.weight) as NEWVAR 572 from sashelp.class a, sashelp.class b 573 where b.height=calculated newvar; NOTE: Table WORK.TEST created, with 23 rows and 6 columns. Your WHERE clause looks invalid since it has the binary operator AND with only one input. Either remove the AND or place some other binary expression before it. proc sql; create table date_naiss as select pers.D_NAISS , sousc.no_pol , coalesce(ctant.ctant_pere,ctant.contrac) as COAL from dossier sousc , dossier2 ctant , personne pers where (1=1) and sousc.contr = calculated COAL and ctant.pers = pers.pers ; quit;     ... View more

Suggestion: your subject line should read something like "create character variable from numeric  values".   Please avoid the use of the word "Format" unless you mean a SAS Format. Format has very specific meaning in SAS related to display of values. Run Proc Contents on any of your data sets and you will see the Format is property of variables.   For instance the Format to display a numeric value with a fixed number of characters and leading zeroes is the Zw.d format where W is the number of characters and D is the number of decimal positions to show. A format of Z4. will display an integer value with 4 characters padding leading zeroes as needed. ... View more