reverse palette colors?

[Fred Oswald]

Hello - I have a simple problem and hopefully a simple solution...that I cannot find.

I've attached the data and R code (the latter is also below) for a histogram generated by the code and data (please load ggplot2 and RColorBrewer packages).

All I want to do is reverse the palette (dark = light and light = dark).  

Hm, as long as I'm being picky(!), I'd also like the light color not to be so light as to be indiscernible (i.e., not white).

Thanks for any help on this!

Fred

==

R Code for histogram:

m <- read.csv("example.csv")

a <- cut(m$var.eff.size,5)

bin <- c(.25, .50, .75)

xi <- quantile(m$e, bin)

mg <- qplot(m$e, data = meta, geom = "histogram", fill=a, main="Frequency Distribution")

mg +  opts(panel.background = theme_rect()) +

geom_vline(xintercept = xi) +

labs(x="e", y="f", fill="e") +

scale_fill_brewer(type="seq")

-
Fred Oswald
Associate Professor
Department of Psychology, Rice University
www.owlnet.rice.edu/~foswald

[David Kahle]

Hi Fred -

The below works easily, although I'm sure there's a way to do it directly from scale_fill_brewer.

m <- read.csv('example.csv')

a <- cut(m$var.eff.size, 5)

bin <- .25*(1:3)

xi <- quantile(m$effect.size, bin)

qplot(effect.size, data = m, geom = 'histogram', 

    fill = a, main = 'Frequency Distribution') +

  geom_vline(xintercept = xi) +

  theme_bw() + labs(x = 'e', y = 'f', fill = 'e') +

  scale_fill_brewer(type = 'seq')

# straight reversal

fills <- rev(brewer.pal(5, 'Blues'))

qplot(effect.size, data = m, geom = 'histogram', 

    fill = a, main = 'Frequency Distribution') +

  geom_vline(xintercept = xi) +

  theme_bw() + labs(x = 'e', y = 'f', fill = 'e') +

  scale_fill_manual(values = fills)

# darker

fills <- rev(brewer.pal(7, 'Blues'))[1:5] 

qplot(effect.size, data = m, geom = 'histogram', 

    fill = a, main = 'Frequency Distribution') +

  geom_vline(xintercept = xi) +

  theme_bw() + labs(x = 'e', y = 'f', fill = 'e') +

  scale_fill_manual(values = fills)

# in general

fills <- c('red','green','blue','purple','orange')

qplot(effect.size, data = m, geom = 'histogram', 

    fill = a, main = 'Frequency Distribution') +

  geom_vline(xintercept = xi) +

  theme_bw() + labs(x = 'e', y = 'f', fill = 'e') +

  scale_fill_manual(values = fills)

Cheers

david.

PS - nice to see ggplot2 has made it into the Psych department at Rice!

--
You received this message because you are subscribed to the ggplot2 mailing list.
Please provide a reproducible example: http://gist.github.com/270442
 
To post: email ggp...@googlegroups.com
To unsubscribe: email ggplot2+u...@googlegroups.com
More options: http://groups.google.com/group/ggplot2
<Oswald - R help.txt><example.csv>

-------------------------------------------------------------

-------------------------------------------------------------

David J. Kahle, Ph.D.

Assistant Professor of Statistical Science

Department of Statistical Science

Baylor University

One Bear Place #97140

Waco, Texas 76798

Office : (254) 710-6102

http://sites.google.com/site/davidkahle/