I believe SAS/Graph Proc Gchart is built to do exactly what you want (including the 'by' groups), without needing any extra processing, etc...   data Have; length home_campu $6; input acadplan $ count1fifth home_campus $; datalines; 155 128 CHSTR 155 9 MIDLO 156 150 MIDLO 156 90 CHSTR 480 99 MIDLO 480 87 CHSTR 212 97 CHSTR 212 64 MIDLO 299 120 CHSTR 299 92 MIDLO 464 135 CHSTR 464 25 MIDLO ; data Have; set Have; label cawardx='Award Level'; cawardx='AAS'; run; axis1 label=(a=90 "Number of Students") order=(0 to 250 by 50) minor=none; axis2 label=('Academic Plan'); pattern1 v=s c=red; pattern2 v=s c=blue; title1 "Chart 3: Annual Five-Year Mean Headroll Enrollment"; proc gchart data=Have; by cawardx; vbar acadplan / sumvar=count1fifth descending subgroup=home_campus raxis=axis1 maxis=axis2 width=5; ... View more

I definitely would not recommend spending a lot of time & effort trying to implement it in SAS, just to see what it looks like. I don't think the results would be useful, and the time wasted would be what I like to call "a fool's errand" 🙂   If there's already code to do it in R, I would recommend first using R to try it out, and see if it provides any useful insight to your data. If the R venn diagrams provide useful insight, then that might justify investing the effort to try to reproduce it in SAS.   ... View more

This example might help:   https://blogs.sas.com/content/graphicallyspeaking/2019/10/09/sas-championship-golf-plotting-the-results/   ... View more

The other suggestions/observations are valid points, but I agree this is the key mystery: "why changing the DPI in SGPLOT seems to slightly change the font size".   Perhaps the developers could check the underlying calculations(?)  @DanH_sas ?   ... View more

Not really knowing what your map looks like, or what other settings are in effect when you run it (eg, what font sizes, or how long the legend text is, etc), I'll just try to give some advice that will get you past the warnings, and at least show a legend (maybe not quite the legend you want ... but at least a starting point).   Perhaps the legend text values are too long to fit in the graph area, or perhaps the text is in a large font which makes it extend outside the graph area. Or, perhaps the legend origin position is causing a problem.   Perhaps if you try something like this instead of the legend statement you're currently using, it will create a legend with very small text font, that will fit in the graph area and not trigger the warning that suppresses the legend (I'm writing these options from memory - they might need some fine-tuning!) ...   goptions htext=5pt; legend1 position=(top left) across=1 mode=share;     ... View more

First, find the latitude/longitude of the stores (you could use Proc Geocode, or some other technique).   You can then easily overlay markers on a map at those lat/long coordinates using Proc SGmap. You could use different colors or shapes for the 2 companies. If you run into the situation where they're in essentially the same location (depending on how far the map is zoomed out ... 20 miles might look about the same on a US map, for example) then be sure to use markers that will still be visible when overlapping, or use transparent colors that will 'mix' well to tell you there's an overlap.   Here's a simple example, showing how to overlay colored markers at lat/long coordinates using SGmap (I use city locations from mapsgfk.uscity to get the lat/longs, but I would recommend using something like Proc Geocode to get street-level lat/long coordinates of the stores):   http://robslink.com/SAS/ods6/colored_dot_map_info.htm   ... View more

Here's one way to plot the facilities on a map, color-coded by 'cat'...   data foo; length cat $10; input id Longitude Latitude Measure Cat Count_Fips_Code; datalines; 1 -85.52972 33.10083 0.6 low 1 3 -85.4075 33.65056 0.5 medium 2 2 -87.165 31.57167 0.8 high 1 ; run; proc sgmap plotdata=foo; openstreetmap; scatter x=longitude y=latitude / group=cat markerattrs=(symbol=circlefilled size=12px) /* More current versions let you specify colors via colormodel=(cxd7191c yellow) */ ; run; ... View more

As PaigeMiller mentioned, it would be interesting to see what you've tried.   That being said, here is one solution ...   proc import datafile="/home/u59782200/datapoints.xlsx" out=datapoints dbms=xlsx; run; proc import datafile="/home/u59782200/balt_city.xlsx" out=balt_city dbms=xlsx; run; proc sgmap mapdata=balt_city plotdata=datapoints; openstreetmap; choromap / mapid=subregion lineattrs=(color=red thickness=2px); scatter x=long y=lat / group=tower markerattrs=(symbol=circlefilled size=12px) /* More current versions let you specify colors via colormodel=(cxd7191c yellow) */ ; run; ... View more

Are you running the code in SAS Studio, Enterprise Guide, DMS SAS, other? Per the 'resolution', have you tried increasing where you're using image_dpi=200 ?   ... View more

Were there any error messages in the SAS log? That should probably point you in the right direction!   You might need to turn on more robust messages in the log to help get more info.  If I'm remembering correctly, something like ...   options source mprint;   ... View more

You could annotate the image on a blank graph page, and then programmatically annotate the names/etc in the desired places (using trial-and-error to guess the desired x/y locations for the text the first time). ... View more

If you only want the date, and the count of the number of obsns for each date, you might try something like ...   proc sql noprint; create table foo as select unique date, count(*) as count from my_data group by date; quit; run;   ... View more

Here's one way to do it ...   data foo; id='001RA'; run; data foo; set foo; id=substr(id,1,length(id)-2); run; proc print data=foo; run; ... View more