Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Roger Harbord <rmharbord@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: RE: x axis forest plot |
| Date | Thu, 7 Oct 2010 18:22:58 +0100 |
You are meta-analysing risk ratios (the "rr" option). -metan- plots ratio effect measures on a logarithmic scale, because that's the sensible thing to do (Galbraith 1988). You are trying to label a ratio of 0. The logarithm of 0 results in a missing value, hence the error message. Try another value instead. xlabel(0.2,1,5) would keep 1 in the centre of the plot. Reference: Galbraith R. A note on graphical presentation of estimated odds ratios from several clinical trials. Statistics in Medicine 1988;7:889-894. DOI:10.1002/sim.4780070807 Regards, Roger. -- Roger Harbord http://www.epi.bris.ac.uk/staff/rharbord.htm On 7 October 2010 14:58, Vincent de Jonge <v.dejonge@erasmusmc.nl> wrote: > one of the commands I used was: > > metan a b c d, label(namevar=Author) by(HQ) random rr xlabel(0,1,5) force > > with this error: > invalid label specifier, : 0 "1" . "0" 0 "1" 1.6094379124341 "5": > > or > metan a b c d, label(namevar=Author) by(HQ) random rr xlabel(0,1,5) > invalid label specifier, : 0 "1" . "0" 0 "1" 1.6094379124341 "5": > > is that of any help? > > on 07-10-2010 14:48 Nick Cox said the following: >> >> -metan- is a user-written command (SSC). >> It usually helps and never hurts if you show exact code that you tried. >> In general with a graphics command you can not omit data points from a >> graph by truncating the axis scale. You must omit data explicitly e.g. by >> using -if-. >> -xaxis()- and -xlabel()- are options, not commands. >> Nick n.j.cox@durham.ac.uk >> Vincent de Jonge >> >> I did a meta-analysis on count data using the db metan command and plotted >> a forest plot. However, to improve readability I would like to limit the x >> axis. This just won't work. I don't know how to do it. Tried commands like >> xaxis, xlabel, etc. but always get an error. Hope someone can get me the >> answer. * * 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/