Hi:   Thanks for posting data. Using your test data and @Reeza's suggestion:  As you can see, I get the correct color coding in columns A and B as you intended with your original code.    I explain in the screen shot that the COMPUTE block for A cannot test the value of C because of the PROC REPORT "left-to-right" rule and a timing issue. At the point in time when PROC REPORT is placing A on the report row, it has not yet placed C on the report row. So the COMPUTE block for A cannot test the value of C. However, you can test the value for C in the COMPUTE block for C. At that point in time, the cell for column A has been placed on the report row, so based on the result of the IF in the COMPUTE block a CALL DEFINE in the COMPUTE block for C can "touch" a variable that appears before C in the COLUMN statement. The only trick is that the variable name you want to change with the call define must be quoted, as I show in my code.     Just a few other comments. HEADLINE, HEADSKIP and WIDTH= are ignored by ODS destinations that would use your STYLE. These are LISTING only options and ignored by ODS, so I removed them from my code. I also changed the colors to lightgreen and lightred to make the numbers in the cells more readable.   Hope this helps, Cynthia ... View more

@JeffMeyers yes, thanks. I saw your code and have being tring to use it as well, but am still not getting it @Jay54 thanks for the link, I think that will be more usefull for networking. I need something to show patients moving from one place(their residence) to the other(to the clinic) ... View more

If you can't find a way to change the font size overall, perhaps you could change it using various 'attrs' size= in the code. See the bolded code in my sample below, for example...   proc sgpanel data=reported_data noautolegend; format date monname1.; panelby county_name / columns=4 rows=5 sort=data novarname noheader spacing=10 border uniscale=all; needle y=cases_this_day_per_100k x=date / displaybaseline=off lineattrs=(color=pink) tip=none; series y=avg x=date / group=county_name lineattrs=(color=blue thickness=1.5 pattern=solid) tip=none; inset county_name / position=top nolabel textattrs=(color=gray33 size=10pt); rowaxis display=(nolabel noline noticks) values=(0 to 200 by 50) valuesdisplay=(' ' '50' '100' '150' '200') valueattrs=(color=gray33) grid offsetmin=0 offsetmax=0; colaxis display=(nolabel noline noticks /*novalues*/) values=('01mar2020'd to '01apr2021'd by month) valueattrs=(color=gray33 size=7pt) grid offsetmin=0 offsetmax=0; run;   ... View more

As a programmer who also works with tumor measurement data my question is: is it necessary to even show the patient ids in the axis? The purpose of the plot is to get an idea of how patients responded to the treatment in a visual, and using the graph to track an individual patient isn't really efficient. I would turn the ticks and labels off for the x axis and just mention how many patients are included. ... View more

Here are the ODS graphics options I use most often. The height and width options may address your issue. I find that imagefmt=PDF results in better looking plots, but the resulting file size can be much larger than imagefmt=PNG, depending on the number of data points.   ODS GRAPHICS ON/reset border = off antialiasmax=1000000 height = 100pct width = 100pct imagefmt=PDF; ... View more

This blog post has complete code on how you make a Forest plot using data that you have available in SAS as a sample data set. You can then replace the data with your own data: https://blogs.sas.com/content/graphicallyspeaking/2017/10/22/tips-tricks-segmented-discrete-axis/ ... View more

Thank you for your help, Jeff. 😄 ... View more