Scatterplot with the 3rd dimension = color?

AFAIK, you can't 'add' two ggplot2 graphs together; the problem in
this case is that the two color scales would clash. If you're willing
to discretize the z values, then you could pull it off. Here's an
example:

d <- data.frame(x = rnorm(100), y = rnorm(100), z = factor(1 +
(rnorm(100) > 0)))
d1 <- data.frame(x = rnorm(100), y = rnorm(100), z = factor(3 +
(rnorm(100) > 0)))
dd <- rbind(d, d1)

In each data frame, I'm assigning two factor levels depending on
whether z > 0 or not. The factor levels are 1, 2 in d and 3, 4 in d1;
when rbinded together, z has four distinct levels. Now call ggplot():

ggplot(dd, aes(x = x, y = y, colour = z)) + geom_point() +
? ?scale_colour_manual(values = c('1' = 'red', '2' = 'blue', '3' = 'green',
? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? '4' = 'yellow'))

This may be coarser than you like, so you could always use the cut()
function to discretize z in each data frame; you'll want to assign the
levels so that they are distinct in the combined data frame. Example:

d3 <- data.frame(x = rnorm(100), y = rnorm(100),
? ? ? ? ? ? ? ? ?z = cut(rnorm(100), breaks = c(-Inf, -0.5, 0.5, Inf),
labels = 1:3))
d4 <- data.frame(x = rnorm(100), y = rnorm(100),
? ? ? ? ? ? ? ? ?z = cut(rnorm(100), breaks = c(-Inf, -0.5, 0.5, Inf),
labels = 4:6))
dd2 <- rbind(d3, d4)

mycols <- c('red', 'maroon', 'blue', 'green', 'cyan', 'yellow')
ggplot(dd2, aes(x = x, y = y, colour = z)) + geom_point() +
? ?scale_colour_manual(breaks = levels(dd2$z),
? ? ? ? ? ? ? ? ? ? ? ?values = mycols)

You can always use the labels = argument of scale_colour_manual() to
assign more evocative legend values, or equivalently, you can assign
the labels in the cut() function within d3 and d4 to those you want in
the legend and leave the plot code as is.

BTW, there is a dedicated ggplot2 list to which you can subscribe
throughhttp://had.co.nz/ggplot2/(look for the ggplot2 mailing list
near the top of the page). The list archives are accessible through
the same link.

HTH,
Dennis









On Thu, Oct 20, 2011 at 12:25 PM, Kerry <kbro... at gmail.com> wrote:

Can someone please help me out with this? The ggplot2 suggestion works
great but I've spent a few days trying to figure out how to plot 2
variables with it and I'm stuck. Here's my example code:

library(ggplot2)
#Here's the 1st plot
x<-rnorm(100)
y<-rnorm(100)
z<-rnorm(100)
d <- data.frame(x,y,z)
dg<-qplot(x,y,colour=z,data=d)
dg + scale_colour_gradient(low="red", high="blue")

#Here's the 2nd plot which will delete the 1st plot above but I'd
like
them to be plotted together
x1<-rnorm(100)
y2<-rnorm(100)
z3<-rnorm(100)
d1 <- data.frame(x1,y1,z1)
dg1 <-qplot(x1,y1,colour=z1,data=d1)
dg1 + scale_colour_gradient(low="green", high="yellow")

I've been trying to get long format working but it just doesn't make
any sense to me.

Thanks,
kb

On Oct 17, 3:10?pm, Kerry <kbro... at gmail.com> wrote:

Yes, the qplot works great, but do you know how to allow for multiple
plots? I want one variable to be plotted say from blue to red and
another say from yellow to green but in the same graph, each having
there own separate legends. I've tried print() and arrange() but no
luck.

Thanks again,
kb

On Oct 2, 10:42?pm, Ben Bolker <bbol... at gmail.com> wrote:

Duncan Murdoch <murdoch.duncan <at> gmail.com> writes:

On 11-10-02 1:11 PM, Kerry wrote:

I have 3 columns of data and want to plot each row as a point in a
scatter plot and want one column to be represented as a color gradient
(e.g. larger ?values being more red). Anyone know the command or
package for this?

It's not a particularly effective display, but here's how to do it. ?Use
rainbow(101) in place of rev(heat.colors(101)) if you like.

x <- rnorm(10)
y <- rnorm(10)
z <- rnorm(10)
colors <- rev(heat.colors(101))
zcolor <- colors[(z - min(z))/diff(range(z))*100 + 1]
plot(x,y,col=zcolor)

? or

d <- data.frame(x,y,z)
library(ggplot2)
qplot(x,y,colour=z,data=d)

? I agree about the "not particularly effective display"
comment, but if you have two continuous predictors and
a continuous response you've got a tough display problem --
your choices are:

? 1. use color, size, or some other graphical characteristic
(pretty far down on the "Cleveland hierarchy")
? 2. use a perspective plot (hard to get the right viewing
angle, often confusing)
? 3. use coplots/small multiples/faceting (requires
discretizing one dimension)