forrest plot

Dear Thomas, thank you very much for the informations, I want just that one!

But how can I do? Can you tell me where can I try the commands to do that 
graphic?

Thank you a lot
Michela

**************************************
Dr.ssa Michela Ballardini
Unit?? di Biostatistica e Sperimentazioni Cliniche
c/o Osp. Morgagni-Pierantoni - Pad. Valsalva
Via Forlanini, 34
47100 Forl??
Tel 0543-731836
Tel/Fax 0543-731612
**************************************


----- Original Message ----- 
From: "Thomas Lumley" <tlumley at u.washington.edu>
To: "Michela Ballardini" <m.ballardini at ior-forli.it>
Cc: <r-help at stat.math.ethz.ch>
Sent: Thursday, October 20, 2005 8:19 PM
Subject: Re: [R] forrest plot



metaplot() will not currently do asymmetric confidence intervals, which
seems to be what you want.  I'm making a revised version based on Paul
Murrell's grid example at
   http://www.stat.auckland.ac.nz/~paul/RGraphics/examples-table.png
which will be more flexible.

  -thomas

Hello
I'm trying to plot hazard risk values using the function metaplot with
the specifications:

metaplot(mn=c(-0.28174,-0.71444,-0.12375,-0.12426,-0.30011,-0.45058,-0.07324),se=c(0.20766,0.42691,0.26366,0.30357,0.31819,0.28636,0.37758),xlab="HR 
and 95%CI",logeffect=T,xaxt="n")

axis(side=1,at=c(0,0.2,0.4,0.6,0.8,1.0,1.2,1.4,1.6,1.8,2.0),labels=c(0,0.2,0.4,0.6,0.8,1.0,1.2,1.4,1.6,1.8,2.0))



However, in the plot the x axis is on a log scale and tends to overextend 
the left end of the axis. How can I transform the x-scale on a linear 
scale with equidistant points?


Thank you very much for your attenction

Mic

**************************************
Dr.ssa Michela Ballardini
Unit?? di Biostatistica e Sperimentazioni Cliniche
c/o Osp. Morgagni-Pierantoni - Pad. Valsalva
Via Forlanini, 34
47100 Forl??
Tel 0543-731836
Tel/Fax 0543-731612
**************************************


[[alternative HTML version deleted]]

Thomas Lumley Assoc. Professor, Biostatistics
tlumley at u.washington.edu University of Washington, Seattle