Hi.  If you are asking how to display that value in scientific notation, use a format ...   data _null_; x = 0.000082634; put x e9.; run;   If you are really asking how to convert numeric a numeric value, it's an odd question since SAS stores all numeric values in the same manner (unless you start playing with numeric value lengths).  It's just  the display of the value that you can alter. ... View more

Hi.  The behavior you describe relies on the RESULTS setting you have specified under the tab TOOLS/OPTIONS/PREFERENCES (see attached).  This adds a GPATH= option to the ODS HTML statement that it generates, and that's what directs your output to the temp folder.   If  you have selected just HTML, you get the behavior you described.  If you have selected LISTING (as I have) you'll get ...   52 filename graphout 'z:\test.png'; 53 54 goptions reset=all device=png gsfname=graphout; 55 56 proc gplot data=sashelp.class; 57 plot height*weight; 58 run; NOTE: 9532 bytes written to z:\test.png. 59 quit;   If you have both HTML and LISTING checked, you get ...   61 filename graphout 'z:\test.png'; 62 63 goptions reset=all device=png gsfname=graphout; 64 65 proc gplot data=sashelp.class; 66 plot height*weight; 67 run; NOTE: Writing HTML Body file: sashtml1.htm NOTE: 9520 bytes written to C:\Users\Mike\AppData\Local\Temp\SAS Temporary Files\_TD1112_MIKENEW_\gplot6.png. NOTE: 9532 bytes written to z:\test.png. 68 quit;         If you don't want to specify ODS LISTING, you can redirect the ODS HTML that SAS generates by adding:    ods html gpath="<your path>";   Before your PROC GPLOT statement. ... View more

Hi, here's some ZIP info (I'll leave the interpolation to someone else).   There's a SAS-supplied data set, SASHELP.ZIPCODE, that has info you can use to find ZIPs associated with counties.  Note, there are ZIPs that cross county boundaries and the ZIPCODE data set only lists one county per ZIP.   proc sort data=sashelp.zipcode (keep=state county zip zip_class) out=new (drop=zip_class); by state county; where missing(zip_class); run;   gives you a data set with STATE, COUNTY, and ZIP  (the WHERE statement eliminates PO BOXES and unique ZIPs that have no geography, as with a ZIP associated with a business).   If you use this ...   proc freq data=new; tables state*county / noprint out=table; run; proc freq data=table; table count; run;   you'll see that about 75% of counties have 10 or fewer ZIPs.  However, 5% have 30+ ZIPs (the maximum is 289 ZIPs in a single county).  Not sure how you will use this data and that will determine how you would organize a lookup file.  One way would be to transpose the data ...   proc transpose data=new out=new ( drop=_: ) prefix=zip; var zip; by state county; run;   and you would have a data set with variables STATE, COUNTY, ZIP1-ZIP289.  There would be a lot of missing data among the ZIPs since only 25% of the STATE/COUNTY combinations have 10+ ZIPs, for example here's observation #1 ...   Obs STATE COUNTY zip1  zip2  zip3 zip4  zip5  zip6  zip7  zip8 zip9 zip10 <more> 1    1        1 36003 36006 36008 36051 36066 36067 36749 .     .     .   <more>   ZIP8 through ZIP289 are missing.     If you try the folowing on the tranpsoed data set, you'll see the STAT/COUNTY with the 289 ZIPs ...   proc print data=new ; var state county zip1-zip10 zip289; where zip289 is not missing; run;   You can do some reading ...   ZIP Code 411:  A Well-Kept SAS® Secret http://www2.sas.com/proceedings/sugi31/143-31.pdf   ZIP Code 411: Decoding SASHELP.ZIPCODE and Other SAS® Maps Online Mysteries http://support.sas.com/resources/papers/proceedings10/219-2010.pdf   http://www.sascommunity.org/wiki/ZIP_Code_411:_Decoding_SASHELP.ZIPCODE_and_Other_SAS%C2%AE_Maps_Online_Mysteries   ... View more

Hi ...   data matrix; array r(5); do i=0 to 20 by 5; do j=1 to 5;   r(j)=i+j; end; output; end; keep r: ; run; data new; array r(5); array v(5); do i=1 to 5; do j=1 to 5;   set matrix point=j;   v(j) = r(i); end;   output; end; stop; keep v: ; run;   data set NEW ... Obs v1 v2 v3 v4 v5 1    1  6 11 16 21 2    2  7 12 17 22 3    3  8 13 18 23 4    4  9 14 19 24 5    5 10 15 20 25   If you wanted to make the transpose data step general, you could try ...   proc sql noprint; select nvar, nlobs into :nvars trimmed, :nobs trimmed from dictionary.tables where libname = 'WORK' and memname = 'MATRIX'; quit;   data new; array r(&nvars); array v(&nobs); do i=1 to &nvars; do j=1 to &nobs;   set matrix point=j;   v(j) = r(i); end;   output; end; stop; keep v: ; run;   ... View more