cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
RW9
  • 11,196 Posts
  • 1,041 Likes Given
  • 982 Solutions
  • 4,011 Likes Received
This widget could not be displayed.

Re: Iterative Macro Variable code

by Diamond | Level 26 in SAS Programming
‎03-19-2014 12:27 PM
‎03-19-2014 12:27 PM
Just to add, you can also use call execute to call your macro if you need to: libname NACO 'sas/data/reference';  %macro NACO (ext=); Proc Sql; Create Table naco as select Select count(*) from naco.NAco_&ext.; quit; %Mend NACO; data _null_;   do i=201302 to 201402;     call execute('%NACO (ext='||strip(put(I,best.))||');');     end; run; ... View more

Re: Iterative Macro Variable code

by Diamond | Level 26 in SAS Programming
‎03-19-2014 12:19 PM
‎03-19-2014 12:19 PM
Here's one way - create your empty table.  Then using call execute generate each insert - note the yyymm is just a number so you could format it, and I have just used sashelp.cars, so you could re-work that for your data specifically): proc sql;   create table NACO   (     MNTH num,     CNT num   ); quit; data _null_;   do i=201302 to 201402;     call execute('proc sql;                               insert into NACO                               set MNTH='||strip(put(i,best.))||',                                   CNT=(select count(1) from sashelp.cars);     /* Note change this to - from NACO_'||strip(put(I,best.)||') - for your datasets */                             quit;');     end; run; ... View more

Re: SAS format - Dollar sign on the right of the amount

by Diamond | Level 26 in SAS Procedures
‎03-19-2014 11:25 AM
‎03-19-2014 11:25 AM
Hi, Try picture formats: Base SAS(R) 9.2 Procedures Guide ... View more

Re: Converting long characters into numerics with decimals

by Diamond | Level 26 in SAS Data Management
‎03-19-2014 06:38 AM
3 Likes
‎03-19-2014 06:38 AM
3 Likes
If you need to sort numerically and that's all, could you do it with two variables, one whole, one part and then use both in the sort: data have;   attrib ct_claim_trans_id format=$200.;   ct_claim_trans_id="842825";output;   ct_claim_trans_id="228390.2086";output;   ct_claim_trans_id="20111212111209.5639";output; run; data want;   set have;   attrib ct_claim_trans_no format=best19.5 inter format=best19.;   if index(ct_claim_trans_id,".")>0 then do;     ct_claim_trans_no=input(substr(ct_claim_trans_id,1,index(ct_claim_trans_id,".")-1),best19.);     inter=input(substr(ct_claim_trans_id,index(ct_claim_trans_id,".")+1),best19.)/10000;   end;   else do;     ct_claim_trans_no=input(strip(ct_claim_trans_id),best.);   end; run; proc sort... by ct_claim_trans_no inter; run; ... View more

Re: Array code

by Diamond | Level 26 in SAS Procedures
‎03-19-2014 06:19 AM
‎03-19-2014 06:19 AM
...   array in{4} (...);   array out{4} (...);   array dtmin{5} (...);   array dtmax{5} (...);   do i=1 to 4;      do z=1 to 5;         if (dtmin{z} <= in{i} <= dtmax{z}) or (dtmin{z} <= out{i} <= dtmax{z}) then flag="Y";      end;   end; ... ... View more

Re: how to calculate the difference for row data

by Diamond | Level 26 in SAS Programming
‎03-19-2014 04:53 AM
‎03-19-2014 04:53 AM
You just need to look at adding if statements as appropriate (marked where I would put them below), logic for that I will leave to you: data want;   set have;      attrib diff format=$10.;   by cust_no;   retain lstamnt lstold lstbal; /* Insert your logic to work out that column, e.g. if something < something else then other=something-something else */    if first.cust_no then do;     lstamnt=amount;     lstold=old;     lstbal=balance;     output;   end;   else do;     output;     cust_cntry="";     datetotal="";     amount=lstamnt-amount;     old=lstold-old;     balance=lstbal-balance;      if balance < 0 then diff="Under";      else if balance=0 then diff="Equal";      else diff="Over";     output;   end; run; ... View more

Re: how to calculate the difference for row data

by Diamond | Level 26 in SAS Programming
‎03-18-2014 01:54 PM
‎03-18-2014 01:54 PM
Hi, Just add an if statement in (note untested as no longer at work): data want;   set have;      attrib diff format=$10.;   by cust_no;   retain lstamnt lstold lstbal;   if first.cust_no then do;     lstamnt=amount;     lstold=old;     lstbal=balance;     output;   end;   else do;     output;     cust_cntry="";     datetotal="";     amount=lstamnt-amount;     old=lstold-old;     balance=lstbal-balance; /****/      if balance < 0 then diff="Under";      else if balance=0 then diff="Equal";      else diff="Over"; /****/     output;   end; run; ... View more

Re: how to calculate the difference for row data

by Diamond | Level 26 in SAS Programming
‎03-18-2014 11:17 AM
‎03-18-2014 11:17 AM
Hi, Well, for two values its real simple with first/last and by: data have;   attrib Cust_No format=8. Cust_cntry Datetotal format=$10. amount old balance format=8.;   infile datalines;   input cust_no cust_cntry $ datetotal $ amount old balance; datalines; 1001 IN 4/30/2012 10000.00 2000.00 8000 1001 IN 4/30/2012 10000.00 2000.00 8000 1002 UK 4/31/2012 15000.00 3000.00 12000 1002 UK 4/31/2012 15000.00 2000.00 13000 ; run; data want;   set have;   by cust_no;   retain lstamnt lstold lstbal;   if first.cust_no then do;     lstamnt=amount;     lstold=old;     lstbal=balance;     output;   end;   else do;     output;     cust_cntry="";     datetotal="";     amount=lstamnt-amount;     old=lstold-old;     balance=lstbal-balance;     output;   end; run; For more rows you just need to have a third if statement to capture not first and not last. ... View more

Re: Difference between itt and safety populations

by Diamond | Level 26 in SAS Health and Life Sciences
‎03-18-2014 09:45 AM
1 Like
‎03-18-2014 09:45 AM
1 Like
Hi, This is not really a SAS specific discussion.  It should be defined in your specs/protocol somewhere what these are.  As for actual definitions - Wikipedia: http://en.wikipedia.org/wiki/Intention_to_treat_analysis So ITT is those randomized based on randomized treatment not what they were actually given. Safety is all randomized with any treatment. Again, might be slightly different depending on your specs. ... View more

Re: SAS Beginner

by Diamond | Level 26 in SAS Programming
‎03-18-2014 04:59 AM
3 Likes
‎03-18-2014 04:59 AM
3 Likes
Hi, This sounds more like a CDISC question rather than SAS, please consult the documents on their website for domains and what each one needs to contain e.g. the implementation guidelines.  You create Adam/SDTM datasets in SAS exactly the way you create any others, the actual logic of it however is CDISC. ... View more

Re: SAS macro parameter message on hover

by Diamond | Level 26 in SAS Programming
‎03-17-2014 11:24 AM
‎03-17-2014 11:24 AM
I think what you are talking is about "auto-completion" which is found in most modern code editors.  Unfortunately SAS's Enhanced Editor is a bit of old beast.  The nearest thing (at least to V9.3) is abbreviations, Tools->Add Abbreviation.  Check out the docs for more info: http://support.sas.com/resources/papers/abbreviations07.pdf ... View more

Re: Create a new character variable!

by Diamond | Level 26 in SAS Programming
‎03-17-2014 06:16 AM
‎03-17-2014 06:16 AM
Just to note RichardinOz, best. will return 8 characters but default so will be truncated by the $5.  Pop a strip() around the putn, or increase the with to $10. ... View more

Re: Slect max

by Diamond | Level 26 in SAS Programming
‎03-14-2014 11:11 AM
‎03-14-2014 11:11 AM
Hi, Like this?  Gets a bit more complicated if you have more than one variable as you will then need to left join the data back on where custid=sustid and max(flag)=flag. data have;   attrib custid format=$10. mr_flag format=best.;   custid=15013; mr_flag=0; output;   custid=15013; mr_flag=1; output; run; proc sql;   create table WANT as   select  distinct           CUSTID,           MAX(MR_FLAG) as FLAG   from   HAVE   group by CUSTID; quit; ... View more

Re: how to use SYSUSERID as observation.

by Diamond | Level 26 in SAS Programming
‎03-14-2014 06:34 AM
‎03-14-2014 06:34 AM
You can refer to it as per any other macro variable: %if "&SYSUSERID." = "Me" %then %do;      something %end; %else %do; ... ... View more

Re: Summary Table with Daily Total and Month to Date Total

by Diamond | Level 26 in SAS Enterprise Guide
‎03-14-2014 06:29 AM
‎03-14-2014 06:29 AM
Well, you can do the below - assuming that your dates are all the same structure: /* Test data */ data have;   attrib dte format=$10. anum format=best.;   dte="Mar 1"; anum=1; output;   dte="Mar 1"; anum=1; output;   dte="Mar 1"; anum=1; output;   dte="Mar 1"; anum=1; output;   dte="Mar 1"; anum=1; output;   dte="Mar 2"; anum=1; output;   dte="Mar 2"; anum=1; output;   dte="Mar 2"; anum=1; output;   dte="Mar 2"; anum=1; output;   dte="Mar 3"; anum=1; output;   dte="Mar 3"; anum=1; output;   dte="Mar 3"; anum=1; output;   dte="Mar 3"; anum=1; output;   dte="Mar 3"; anum=1; output;   dte="Mar 3"; anum=1; output;   dte="Mar 3"; anum=1; output; ; run; proc sql;   create table WANT as   select  DAILY.DTE,           DAILY.DAILY_SUM,           (select SUM(THIS.ANUM)            from   HAVE THIS            where  substr(DAILY.DTE,1,3)=substr(THIS.DTE,1,3)             and   input(strip(substr(THIS.DTE,4)),best.)<=input(strip(substr(DAILY.DTE,4)),best.)           ) as MTD_TOTAL   from    (             select  distinct                     DTE,                     SUM(ANUM) as DAILY_SUM             from    HAVE             group by DTE           ) DAILY; quit; ... View more
Contact Me
Group Hubs for RW9
View All
Likes given to
Quick links