Hi All,   When I ask for plots in proc phreg and use the baseline covariates statement to ask for a direct adjusted plot, the title "Direct Adjusted Survivor Functions" appears. How do I get rid of this? I have tried modifying the title by using ods trace to find the template for the survival graph and modify the proc template code. However, the template 'Stat.Phreg.Graphics.Survival' does not have a title with "Direct Adjusted Survivor Functions." Is this the correct template for this graph? If so, how can I modify the code below to get rid of the title?   Thanks! Sarah   proc template; source Stat.Phreg.Graphics.Survival; define statgraph Stat.PHReg.Graphics.Survival; dynamic title1 title2 title3 title4 xviewMin xviewMax group groupIndex groupName plotCL plotHPD transparency piecewise _byline_ _bytitle_ _byfootnote_; BeginGraph; entrytitle TITLE1 TITLE3; entrytitle TITLE2 TITLE4 / textattrs=GRAPHVALUETEXT; layout overlay / xaxisopts=(linearopts=(viewmin=XVIEWMIN viewmax= XVIEWMAX)) yaxisopts=(label="Survival Probability" shortlabel= "Survival" linearopts=(viewmin=0 viewmax=1 tickvaluelist=(0 .2 .4 .6 .8 1.0))); if (PLOTCL) bandplot LimitLower=LOWERSURVIVAL LimitUpper=UPPERSURVIVAL x= TIME / group=GROUP index=GROUPINDEX modelname="Survival" datatransparency=transparency; endif; if (PLOTHPD) bandplot LimitLower=LOWERHPDSURVIVAL LimitUpper= UPPERHPDSURVIVAL x=TIME / group=GROUP index=GROUPINDEX modelname="Survival" datatransparency=transparency; endif; if (EXISTS(PIECEWISE)) seriesplot y=SURVIVAL x=TIME / group=GROUP index=GROUPINDEX name="Survival"; else stepplot y=SURVIVAL x=TIME / group=GROUP index=GROUPINDEX name="Survival"; endif; if (EXISTS(GROUP)) discretelegend "Survival" / location=outside title=GROUPNAME; endif; endlayout; if (_BYTITLE_) entrytitle _BYLINE_ / textattrs=GRAPHVALUETEXT; else if (_BYFOOTNOTE_) entryfootnote halign=left _BYLINE_; endif; endif; EndGraph; end; run; ... View more

Dear All,   I have a dataset with hundreds of dates  and doses of medications. DATE1 has a corresponding DOSE1 and so on. I have passed the values of the dates into new variables based on each person's start date. I would also like the dose values that correspond to the dates to get passed to new variables. For each person, I want the dates that occur on or after the STARTDATE to be passed into new variables called NEW_DATES1, NEW_DATES2, etc.     HAVE ID STARTDATE DATE1 DOSE1 DATE2 DOSE2 DATE3 DOSE3 1 10/25/2000 1/2/1999 20 11/2/2000 20 3/5/2001 50 2 5/16/2005 6/14/2001 15 7/5/2002 100 5/25/2006 10 3 1/18/2004 1/19/2004 25 2/24/2005 20 8/16/2005 25   Want ID STARTDATE NEW_DATE1 NEW_DOSE1 NEW_DATE2 NEW_DOSE2 NEW_DATE3 NEW_DOSE3 1 10/25/2000 11/2/2000 20 3/5/2001 50 . . 2 5/16/2005 5/25/2006 10 . . . . 3 1/18/2004 1/19/2004 25 2/24/2005 20 8/16/2005 25   I was able to pass the old dates to the new dates, but need help with the doses:   array dx (*) date1 - date800; array dt (*) new_date1 - new_date800 do j = 1 to dim(dx) UNTIL (dx(j)=. );        if dx(j) >= startdate then do;           do k=1 to 614-j;              dt(k) = dx(j);              j= j+1;           end;        end; end;   Thanks! Sarah ... View more

Hi All,   I'm trying to use the datepart and timepart functions to extract the date and times from 1 variable called admitdatetime. For example if the value of admitdatetime is 14MAR06:12:15:00 then date=14MAR06 and time=12:15. Further, I would like to create a variable called day that tells me the day of the week the person arrived.    Data want; set have ; informat admitdatetime DATETIME16.; format admitdatetime DATETIME16.; date_arr= datepart(admitdatetime); time_arr=timepart(admitdatetime); day=weekday(date_arr); run;   However, I get missing values for all of these new variables. The admitdatetime is numeric, informat is DATETIME16. so the datepart/timepart function should work.    Thanks, Sarah ... View more

Hi All,   I have calculated the time in hours (variable name timehrs) between two datetime variables. I have converted time in hours to time in minutes (timemin). Both of these variables have time formats on them. I'd like to use timemin as a numeric variable to use in bivariate and multivariable calculations. I cannot figure out how to use the actual time in either hours or minutes instead of SAS time. For instance, if the value of timemins should be 1632 minutes, the value being used in calculations is 5875200, which is SAS time. How do I convert timemins and timehrs to real minutes and hours?   My code: Data have; set a; by Study_Number; btime=dhms(date_b, 0,0,time_b); ctime=dhms(date_c, 0,0,time_c); timehrs=ctime-btime; timemin=(timehrs*60); format btime ctime datetime18. timemin timehrs time4.; run;  ... View more