Could you not just annonymize the data you have, so anything identifiable would be replaced with some code + random number? ... View more

Another method: proc sql;   insert into BASE_TABLE (SYS_CD,SYS_DESC)   select  *   from    DATA_TABLE; quit; ... View more

Hi, Sorry, I am not following your question: " I want only a single graph to refresh every 5 seconds with the latest iterated event".  Graphs are output, there is no 'refresh', you would need to run the program again.  Maybe you want to output each graph to a separate file, e.g. graph_output_time_0, graph_output_time_5, ...? ... View more

Have you tried setting the AUTOFIT_HEIGHT option, and ROW_HEIGHTS? I am not sure there is such an options as height in templates, at least I can't find it in the templates I was looking at? ... View more

To add, I concur with Reeza, 6000 variables is not plausible, even turning the table round so you have 6000 rows and 711 variables would drastically increase your performance.  If you have that many variables then the data should be normalized, e.g. Date          Col          Result JUN14       1             0 JUN14        2            1 JUN14       3             1 MAY14      1              0 MAY14      2              0 ... Whilst working with that many variables may work in SAS, its not a good idea.  Also, depending on your data, try summarizing, grouping, etc to shrink the data up. ... View more

Hi, There are several ways to do this, first one that springs to mind for me is a retain: data want;      set have;      retain lstval;      if base ne . then lstval=base;      else base=lstval; run; There are other suggestions if you search the forums. ... View more

Hi, I wouldn't expect there to be any difference in the operation of the libname syntax against the connect in the proc sql as the actual connection would be a tiny amount of time.  I would suggest that you have a look at the data on the database and see if there are ways to cut down the amount of data to be output, firstly, do you really want all the variables to come out, if not remove the *.  How many records are you talking about with the where ASOFDATE='15-AUG-2014', restriction?  Maybe try parting it out into smaller subgroups.  It could also be other things, network speed, where files are being written to, amount of ram/processor etc. Have you tried running the query in Toad or similar to see the operation there?  Also consider writing the data directly out to CSV from the database, then import to SAS, maybe quicker. ... View more

Hi, Something like: data have; id=1;result="010306"; output; id=2;result="0204"; output; id=3;result="01"; output; id=4;result="0502"; output; id=5;result="0205"; output; id=6;result="01040507"; output; run; data want (drop=i);   set have;   array result_out{7};   do i=1 to 7;     if index(result,"0"||strip(put(i,best.)))>0 then result_out{i}=i;   end; run;     ... View more

Added, just seen LinusH's response and complete concur that transposing so that your data looks like: COL1     0 COL2     0 COL3     0 COL1     1 ... Would be a good idea. Hi, Maybe something like: data have;   col1=0; col2=0; col3=0; output;   col1=1; col2=0; col3=2; output;   col1=1; col2=0; col3=3; output; run; proc sql;   select  case  when sum(COL1)=0 then "COL1"                 when sum(COL2)=0 then "COL2"                 when sum(COL3)=0 then "COL3"                 else "" end   into    :VARNOM   from    HAVE; quit; %put &VARNOM.; ... View more

Hi, Use the out= options to store the output from your proc tabulate.  Drop rows you don't want, then proc report the remaining dataset. ... View more