About Reeza

Reeza · ‎09-27-2011

I'm creating a map using GMAP and the variable being plotted is continuous and then formatted using a proc format. Some of the ranges in the format are not within the dataset and usually you insert fake values to get around this as per the example I've linked to. However this isn't working for me, and I was wondering if there was another way to get all the legend values to show. ie legend values are: 0-9 10-19 20-29 30-39 40-49 but what ends up showing on the map is 0-9 10-19 40-49 because there is no 20-39. I'd like to force those values in the legend. http://communities.sas.com/message/33948#33948

Reeza · ‎09-26-2011

From what I'm reading cccording to the documentation and formula's there is NO correction for tied values. You can test this by running for some data that has no ties and then adding in some tied values around the mid range (this shouldn't change the CL but will change the median) from what i'm seeing. Hopefully that makes some sense.

Reeza · ‎09-23-2011

Docs: http://support.sas.com/documentation/cdl/en/statug/63962/HTML/default/viewer.htm#statug_npar1way_sect013.htm Does that answer your question?

Reeza · ‎09-22-2011

Have you looked at the odds ratio with the oddsratio statement?

Reeza · ‎09-21-2011

The X11 was available from the Apple website at one point as well, but definitely on your install disks. And yes Terminal will connect you, default installed unter utilities last time I checked.

Reeza · ‎09-20-2011

Why not limit to only cases that are in both models for consistency? You can also compare the model estimates and RMSE for the model and then without the observations it would lose by this method to see the effect.

Reeza · ‎09-20-2011

Try adding the sparse option, but then you'll probably need to modify your where clause as well. proc freq data=have noprint; tables role*answer/missing out=sample outpct sparse; run;

Reeza · ‎09-20-2011

So how are you assigning the names? Where do they come from? You might want to google CNTLIN option with proc format and see if that helps.

Reeza · ‎09-20-2011

There's many ways. Here's one two step procedure. 1. Run a proc freq with the missing option proc freq data=have noprint; tables role*answer/missing out=sample outpct; run; 2. Process data to format you need data want; set sample; where answer ne .; response_rate=put(count, 8.)||"("||put(pct_row/100, percent8.)||")"; run; That should help you get started.

Reeza · ‎09-19-2011

Assuming enquiry_date is actually a date field, I think SAS may not be recognizing dataend as a date variable. try %LET dataend = '19sep2011'd; and rerun.

Reeza · ‎09-19-2011

Two options I can see, one is use ODS TEXT= to insert titles. Second is to insert titles, print a blank line and then your output. Trick Is I'm not sure how to 'print' a blank line without a title, though I'm sure you could...proc report is my guess. The following uses the ods text= option (no errors here). Neither solution is perfect but I think its closer to what you were looking for. ods tagsets.msoffice2k_x file="c:\temp\panels.xls" options(panelcols="1" embedded_titles='yes' gridlines='no') style=normal; ods tagsets.msoffice2k_x text= 'This is my main title'; ods tagsets.msoffice2k_x text= 'This is my second title'; ods tagsets.msoffice2k_x text= 'This is my third title'; ods tagsets.msoffice2k_x options(panelcols="3") ; proc print data=sashelp.prdsale(obs=10); var actual predict country region; where country="CANADA"; title "Canada Sales"; run; proc print data=sashelp.prdsale(obs=10); var actual predict country region; where country="U.S.A."; title "USA Sales"; run; proc print data=sashelp.prdsale(obs=10); var actual predict country region; where country="GERMANY"; title "Germany Sales"; run; ods tagsets.msoffice2k_x options(panelcols="2") ; proc print data=sashelp.prdsale(obs=20); where region="EAST"; title "East Sales"; run; proc print data=sashelp.prdsale(obs=20); where region="WEST"; title "West Sales"; run; ods tagsets.msoffice2k_x close; If you want to try the second add this in to the beginning instead of the ods text statements: title 'This is my main title'; title2 'This is my second title'; title3 'This is my third title'; data blank_line; test=''; run; proc report data=blank_line nowd; columns test; define test/display ''; run;

Reeza · ‎09-19-2011

If you're using ODS RTF you can use the option startpage=never or startpage=no.

Reeza · ‎09-19-2011

I think NickR solution will work if you drop address2 in the by statement and sort statements, but then you might have diff apt numbers. But your examples also don't cover this scenario. You can hardcode the comparison in for each variable but I also think you might need to provide more info, I'm assuming that you you're looking to add in the apt number but not sure. A little more explanation of the logic would help.

Reeza · ‎09-19-2011

If they're all in the same library, scan the dictionary tables for names that fit your criteria using a SQL statement and then use that macro variable in your set statement. You probably need to play with the cases to match a bit, but it should get you started. UNTESTED %macro merge_multiple; *get list of names; proc sql; select into : table_list from dictionary.tables where libname='WORK' and memname like 'DATA%'; quit; data want; set &table_list; run; %mend merge_multiple;

Reeza · ‎09-19-2011

Add a 1 in to the dataset for each row and then use proc transpose. data want1; set have; value=1; run; proc transpose data=want1 out=want2; by id date; var value; id vaccine; run;