[R] Barplot coloring question
vasishth at coli.uni-sb.de
Fri Jul 26 14:18:03 CEST 2002
On Thu, 25 Jul 2002, Marc Schwartz wrote in response:
> The "list" of colors that you have specified for colorlist contains only
> one entry.
> If you specify:
> colorlist <- c("red", "blue", ...)
> You bars or their stacked segments will cycle through the colors,
> depending upon whether beside is TRUE or FALSE, respectively.
> Just be sure to specify one color for each group in your data.
Right. I had tried that earlier but that alone didn't work; sorry, I
should have mentioned that. But having beside as TRUE works (with a list
of colours). I wanted to keep beside = FALSE because I wanted to control
the space between bars, as I decribe below with toy examples.
In the first kind of barplot I want, I have data like this:
A A B B C C
0 5 10 15 20 70 90
I can plot it using code like this:
data2 <- as.matrix(read.table("test2.data"))
colorlist2 <- c("orange","limegreen")
width = c(1,.55),
space = c(2.75,.25),
axes = F
I want to be able to manipulate the spaces between the bars pairwise: the
contiguous AA should be closer to each other than the contiguous AB. I
thought that a vector specification for space would do this, because this
works for the plot of a slightly different dataset that I have (below).
But it doesn't in the above case.
Here's an example where "space" DOES work as (I) expected:
The data set:
A A B B C C
1 5 10 15 20 70 90
2 20 15 24 56 34 70
3 40 54 23 67 12 78
data3 <- as.matrix(read.table("test3.data"))
colorlist3 <- c("orange","lightblue","darkblue")
xlim = c(1,6),
width = c(.5,.35),
space = c(2.75,.25),
axes = F
As expected, "colorlist" in both examples cycles through the list of
colors, but when beside = TRUE it uses each color for each column in the
data, and when beside = FALSE the list of colors is cycled through for
each bar's components. In the *first* case, I'm trying to get the two
colors to alternate (which works only when beside = TRUE), but for some
reason I then lose control over "space".
Any suggestions would be greatly appreciated.
> Date: Thu, 25 Jul 2002 13:29:06 -0500
> From: Marc Schwartz <mschwartz at medanalytics.com>
> To: 'Shravan Vasishth' <vasishth at CoLi.Uni-SB.DE>, r-help at stat.math.ethz.ch
> Subject: RE: [R] Barplot coloring question
> > Hi all,
> > I have the following dataset, call it test.data (30 columns, and one
> > named "0"):
> > ADVP ADVP AP AP CONJ CONJ CP CP DU DU INF INF
> > MWU NP NP PP PP PPRT PPRT REL REL SMN SMN SSB SSB SV1
> > SV1 TI TI
> > 0 96.85 2.05 89.07 2.54 70.91 2.37 94.92 3.46 82.31 11.33 40.96 2.25
> > 3.43 90.77 17.63 86.60 10.78 60.27 1.32 93.27 0.97 77.60 12.46 64.98
> > 88.54 3.70 24.33 0.28
> > I'm trying to plot these data points so that any pair of identical
> > (e.g., ADVP) are plotted closer to each other than any neighbouring
> > This much is easy to do:
> > data<-as.matrix(read.table("test.data"))
> > colorlist <- c("red")
> > # build basic barplot:
> > barplot(data,
> > beside=FALSE, # don't put bars beside each other
> > col=colorlist, # list of colors
> > xlim = c(1,40), # x-axis limits
> > width = .65, # width of bar plots
> > ylim=c(0, 140), # y-axis limits
> > space = c(1,.5), # space between bars, and between bar-pairs
> > axes = F # don't plot y-axis yet
> > )
> > This gives a graph with the same color for each bar.
> > The problem is that I want the first instance of a pair of identical
> > columns to have a certain colour, and the other one a different one.
> > the bars should really have an alternating pattern like red green red
> > green... and so on, not just one colour.
> > For the above dataset I could set beside=TRUE and transform the
> dataset to
> > two rows and 15 columns rather than the current 1x30, so something
> > ADVP (etc.)
> > 0a 96.85
> > 0b 2.05
> > but I will eventually need to plot a matrix like
> > ADVP ADVP (etc.)
> > 1 30 40
> > 2 20 20
> > 3 10 5
> > and there I will want beside=FALSE because I want the 1, 2, 3 rows to
> > appear *within* a bar, but I still want pairs of bars for ADVP.
> > This is why I don't want to set beside=FALSE.
> > I'll be very grateful for any suggestions.
Dr. Shravan Vasishth Phone: +49 (681) 302 4504
Computational Linguistics, Universität des Saarlandes, Postfach 15 11 50
D-66041 Saarbrücken, Germany http://www.coli.uni-sb.de/~vasishth
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch
More information about the R-help