I have a basic line graph (created through proc sgplot) that I am formatting for a report. Have always hated the grey/light tick marks and axis lines and want to make them both black/bold. I cannot seem to find how to do it but I imagine is has to be out of the proc sgplot statement?   Thank you for any help you can provide! ... View more

I have a dataset that looks like this: Case_ind Dateofvisit Personid 0 3/20/2016 1 0 3/24/2016 1 1 4/3/2016 1 0 4/3/2016 1 0 4/4/2016 1 0 4/20/2016 1 0 6/20/2017 1 0 6/31/2017 1 0 7/5/2017 1 0 7/9/2017 1 1 7/10/2017 1 0 7/11/2017 1 0 7/12/2017 1 0 7/20/2017 1 0 1/16/2016 2 0 1/16/2016 2 1 1/17/2016 2 0 1/18/2016 2 0 1/25/2016 2 1 5/9/2016 3 0 9/16/2016 3 0 9/18/2016 3 1 9/20/2016 3 0 9/27/2016 3   It is a dataset that has doctor visits 3 weeks before and 3 weeks after for a person that was diagnosed (case_ind=1) for a condition at the doctor's office. As you can see, this could happen multiple times for one person. I want to group the visits into *visit before diagnosis* and *visit after diagnosis* so I can look into what happened in a health care setting before and after that person was diagnosed.   This would presumably look like this: Case_ind Dateofvisit Personid Place 0 3/20/2016 1 1 0 3/24/2016 1 1 1 4/3/2016 1 0 0 4/3/2016 1 2 0 4/4/2016 1 2 0 4/20/2016 1 2 0 6/20/2017 1 1 0 6/31/2017 1 1 0 7/5/2017 1 1 0 7/9/2017 1 1 1 7/10/2017 1 0 0 7/11/2017 1 2 0 7/12/2017 1 2 0 7/20/2017 1 2 0 1/16/2016 2 1 0 1/16/2016 2 1 1 1/17/2016 2 0 0 1/18/2016 2 2 0 1/25/2016 2 2 1 5/9/2016 3 0 0 9/16/2016 3 1 0 9/18/2016 3 1 1 9/20/2016 3 0 0 9/27/2016 3 2     I created the following code after messing around for awhile: data trial_final_data; set pre_final_data; by personid ; if first.personid then place=1; else if dif(visitdate)>7 then place=1; if case_ind>0 then place+1; run; data trial_final_data1; set trial_final_data; if case_ind>0 then place=0; run;   The code almost worked but there are random errors after the first data step. PersonID 3 would have something that looked like this: 1 5/9/2016 3 2 0 9/16/2016 3 2 0 9/18/2016 3 2 1 9/20/2016 3 3 0 9/27/2016 3 3   You'll notive that the 2nd observation should be a 1 for the place...but it is a 2. This only happened with like 7 out of 15k people.   Does anyone have an idea why this is going on or another approach to get the end result I would like?   Thank you so much! ... View more

Dealing mainly with the following varaibles:   Case_ind Dateofvisit PersonID   What i have done is define with other variables whether or not a doctor visit was related to my condition of interest (Case_ind 0/1). Now i need to pull out all other visits for the same person that are within 3 weeks before or after the visit for the condititon of interest (case_ind=1). I know how to do this with only 3 weeks after by using a retain statement in a data set by personID, dateofvisit, and case_ind but I am not sure how to grab the ones 3 weeks before?   Any help would be greatly appreciated!! ... View more

I am submitting my figure to a journal that requires a TIFF that is 1200 dpi. I keep trying a bunch of differnt ways but I always get the same message: WARNING: A very large output size of (X, Y) is in effect. This could make Java VM run out of memory and result in some Java exceptions. You should reduce the output size or DPI settings. ERROR: Java virtual machine exception. java.lang.OutOfMemoryError: Java heap space.   I assume that 1200 DPI might be too big for SAS?   My latest attempt was with the following code: goptions device=tiff; ods listing gpath='C:\My files\tte2.tiff' ; goptions reset=all device=tiffp xpixels=6000 xmax=6in ypixels=7200 ymax=36in; proc sgplot data=pred1 ; (bunch of code); run;   I have fairly little experience with ods complexities so any advice/explanations would be helpful! ... View more