R help - Jun 2009 - forestplot function in rmeta package

Hi,
It is known that the area of square plotted by forestplot is proportional to the
studies' weights. But since there is not weight parameter in forestplot
function, how does this function proportion the area of each square based on the
related study's weight?

Look forward to your reply,

Carol



      
	[[alternative HTML version deleted]]

At 08:38 08/06/2009, carol white wrote:
>Hi,
>It is known that the area of square plotted by forestplot is 
>proportional to the studies' weights. But since there is not weight 
>parameter in forestplot function, how does this function proportion 
>the area of each square based on the related study's weight?
Carol, you can examine the source code of forestplot to find out how 
it performs all its calculations.
It calls drawNormalCI to draw the study lines
drawNormalCI has a parameter size
somewhere near the middle of the code you can see how it calculates 
info which it supplies as size to drawNormalCI

If you do not like how it does it you can make your own copy of 
forestplot and change it.
>Look forward to your reply,
>
>Carol
>
>
>
>
>         [[alternative HTML version deleted]]
Michael Dewey
http://www.aghmed.fsnet.co.uk