Hi. I am trying to make a simple line series with multiple lines where each line represents a different group. I would like each group line to be a specific color based on a binary indicator variable. In other words, if group A and B are indicator=0 I want them blue (for example), and if groups C, D, and E are indicator=1 I want them red.   How can I do that?  proc sgplot data=mydata; series y=outcome x=time / group=type; run; ... View more

Hi. I have two data sets with dates I am trying to combine based upon a range of days surrounding the dates (i.e., +/- 7 days).    The first dataset has an id, reason for evaluation, injured body part and the date of evaluation.      data have1; input id $ reason $ bodypart $ evaldate :date9.; format evaldate date9.; datalines; S001 injury ankle 01JAN2020 S001 injury knee 17APR2020 S002 injury hand 04JUL2020 S004 injury shoulder 31JUL2020 S005 injury back 01SEP2020 ; RUN;     The second data set has a matching id, a reason and bodypart variable (which may or may not match with the first data set), and then a start and end date representing work days lost.    data have2; input id $ reason $ bodypart $ startdate :date9. enddate :date9.; format startdate date9. enddate :date9.; datalines; S001 injury leg 01JAN2020 31JAN2020 S001 injury knee 14APR2020 21APR2020 S002 injury wrist 07JUL2020 09JUL2020 S003 injury foot 23SEP2021 31OCT2021 S004 injury shoulder 07JUL2020 09JUL2020 S005 injury neck 30SEP2020 01OCT2020 ; RUN; How can I match on a date range (+/- 7 days, for example)? This is necessary since the evaldate and startdate do not always line up. I would want all observations to appear in the combined dataset, but if they don't fall within the 7 days I would want them to be flagged as not having a match and not be merged. I would also want to know the count of lost days (dayslost=intck('day',startdate,enddate+1)) which I'm assuming I would need to run in a separate datastep. I ultimately want to end up with something like the following:  data want; input id $ reason $ bodypart $ evaldate :date9. reason2 $ bodypart2$ startdate :date9. enddate :date9. dayslost match; format evaldate date9. startdate date9. enddate date9.; datalines; S001 injury ankle 01JAN2020 injury leg 01JAN2020 31JAN2020 31 1 S001 injury knee 17APR2020 injury knee 14APR2020 21APR2020 8 1 S002 injury hand 04JUL2020 injury wrist 07JUL2020 09JUL2020 3 1 S003 . . . injury foot 23SEP2021 31OCT2021 39 0 S004 injury shoulder 31JUL2020 . . . . . 0 S004 . . . injury shoulder 07JUL2020 09JUL2020 3 0 S005 injury back 01SEP2020 . . . . . 0 S005 . . . injury neck 30SEP2020 01OCT2020 2 0 ; run; Thanks in advance!   ... View more

Hi. I'm starting to do a poisson regression with proc genmod and wanted to first check that I have my data summarized correctly. I assumed that I could test this with the following crude model without covariates and compare the estimate with a by-hand calculation of outcome/person-time.  proc genmod data=summarized; MODEL outcome= /dist=poisson link=log offset=logdenomcount; estimate "Crude" intercept 1; run; However,  the estimate and my by-hand rate are not the same. Is my assumption these should match incorrect?  ... View more

I'm having trouble figuring out how to create two new variables that depend on the most recent date. Let's say I have ID, dx (diagnosis), a prescribed med, and a start date.  data have; input id dx $ med $ startdate :date9.; format startdate date9.; datalines; 1 a a 01JAN2020 1 a a 04APR2020 1 a a 21APR2020 1 a a 01JUL2020 2 a a 01FEB2020 2 a b 01JUL2020 2 a a 07JUL2020 3 b b 01JAN2020 3 b a 04APR2020 3 b a 01MAY2020 3 b a 15JUN2020 3 b a 01SEP2020 3 b b 02SEP2020 ; RUN; For any id, when there is > 60 day gap from the preceding date, given the same dx and med, I want to call this is a "new" prescription. For the same dx and med <60 days from the most recent date I want to call it a "refill".  data want; input id dx $ med $ startdate :date9. refill $ new $; format startdate date9.; datalines; 1 a a 01JAN2020 0 0 1 a a 04APR2020 0 1 1 a a 21APR2020 1 0 1 a a 01JUL2020 0 1 2 a a 01FEB2020 0 0 2 a a 07JUL2020 0 1 2 a b 01JUL2020 0 0 3 b a 04APR2020 0 0 3 b a 01MAY2020 1 0 3 b a 15JUN2020 1 0 3 b a 01SEP2020 0 1 3 b b 01JAN2020 0 0 3 b b 02SEP2020 0 1 ; RUN; The code I've been trying to use is below.  proc sort data=have out=havesort; by id dx med startdate; run; data want; set havesort; by id dx med; IF FIRST.med THEN DO; TEMP=startdate; refill=0; new=0; END; RETAIN TEMP refill new; IF startdate>TEMP + 60 THEN new+1; IF TEMP<startdate<=TEMP + 60 then refill+1; TEMP=startdate; drop temp; run; However, as you can see, this doesn't reset the 60 day counter to the most recent startdate. How can I do that to get what I want?    Let me know if that does not make sense. Thank you always for the help!       ... View more