Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Charles Lindsey <clindsey@stata.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: Re: st: sts graph: different results with -atrisk- vs -risktable- option |
Date | Fri, 15 Apr 2011 12:31:21 -0500 |
Alex Gamma <alex.gamma@uzh.ch> separately used the -atrisk- and -risktable- options in two -sts graph- calls with data that contained multiple records per subject. There was a large discrepancy in the results from the two calls. The graphed -atrisk- numbers were significantly larger than the -risktable- numbers. Alex compared with -sts list- as well, and found these numbers matched the -risktable- numbers. Some of the -atrisk- numbers actually exceeded the total number of subjects in Alex's dataset. Joerg Luedicke <joerg.luedicke@gmail.com> investigated the issue with -atrisk- and -risktable- that Alex discovered. He also found that some of the -atrisk- numbers exceeded the total number of subjects in the data. This is a bug in -sts graph, atrisk- that can arise in multiple record survival data. We are investigating and will provide a solution in a future ado update. --Charles Lindsey * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/