I don't quite understand your R code (I think your random numbers make
the graph very hard to read!), but I have also written an example for
the interaction effect:
## 400 bubbles to show the interaction of x and z
sq = seq(0, 10, length = 20)
x = rep(sq, 20)
z = rep(sq, each = 20)
y = c(outer(sq, sq, function(x, z) 10 + 1 * x + 0.5 *
z + 0.5 * x * z + runif(1)))
symbols(x, z, y, bg = rgb(1, 0, 0, 0.1), fg = "yellow")
## observe the bubbles from the x-axis or y-axis
Regards,
Yihui
--
Yihui Xie <xieyihui at gmail.com>
Phone: +86-(0)10-82509086 Fax: +86-(0)10-82509086
Mobile: +86-15810805877
Homepage: http://www.yihui.name
School of Statistics, Room 1037, Mingde Main Building,
Renmin University of China, Beijing, 100872, China
On Thu, Sep 18, 2008 at 11:18 AM, Wen Gu <edwardg22486 at hotmail.com> wrote:
hey, you suggestion worked great last time. I was able to create a graph through simulation. However, now I am having trouble manipulating the simulated effects of the graph. The code I attached has simulated data values for all 3 values, and is plotting a graph showing an interaction with all 3 variables. I am trying to create extreme effects by having the circle size increase, with increased X value, while holding the other variable somewhat constant (and then do the other). Yet, when I try to change the mean of these values, the scales merely adjusts to account for the increased means.
Date: Sat, 13 Sep 2008 11:58:28 +0800 From: xieyihui at gmail.com To: edwardg22486 at hotmail.com Subject: Re: [R] bubble(circle) plot help. CC: r-help at r-project.org The function symbols() is in the package 'graphics', so you don't need any add-on packages. For example: # similar code to the example you mentioned set.seed(1) symbols(x = rnorm(50), y = rnorm(50), circles = abs(rnorm(50)), main = 'Circles Plot') Yihui