I have a set of data that can be summarized as in the example below. Some values and two other columns that represent some categorical values.

library(ggplot2)
set.seed(123)
sdata=data.frame( aicat=sample(letters[1:4], 50, replace=TRUE), site=sample(letters[5:6], 50, replace=TRUE), vals=sample(1:1000, 50) )

head(sdata)
  aicat site vals
1     c    f  409
2     c    e  308
3     c    e  278
4     b    e   89
5     c    e  928
6     b    f  537

I want to plot this data with boxplot, where on x I have the aicat variable and where I also group the data by site.
To do that I did the following:

ggplot(sdata, aes(x=aicat, y=vals, fills=site, color=site))+
  geom_boxplot()+
  scale_fill_manual()

That produces the following image:

The problem is the following:
how can I assign to each of these boxplots the following list of colors?

values=c("brown1", "brown4", "gold", "gold4", "dodgerblue", "dodgerblue4", "gray70", "gray40")

I would like that the boxplots in "a" are brown1 (the leftmost) and brown4 (the rightmost),
the one in b gold (the leftmost) and gold4 (the rightmost), and so on. I also need to keep this structure of the plot, so solutions with facet_wrap are not what I am looking for.

Any advice?

>Solution :

You can specify manually this colors in fill parameter out of the aes()

 geom_boxplot(fill = c("brown1", "brown4", "gold", "gold4", "dodgerblue", "dodgerblue4", "gray70", "gray40"))