Extracting the colors assigned by ggplot

[Ashim Kapoor]

Dear All,

Here is a reprex :

x <- as.factor(rep(1:10,10))
y <- rnorm(100)
z <- as.factor(sample(x,100))

ggplot(data = data.frame(x,y,z), aes(x=x,y=y,color = z,shape = z,fill = z)) + geom_point()

I want to list out which color is assigned to which level of the factor z.

How can I do this ?

Best Regards,

Ashim

[Ron Crump]

Dear Ashim,

>Here is a reprex :

Thanks.

>x <- as.factor(rep(1:10,10))
>y <- rnorm(100)
>z <- as.factor(sample(x,100))
>
>ggplot(data = data.frame(x,y,z), aes(x=x,y=y,color = z,shape = z,fill =
>z))
>+ geom_point()
>
>I want to list out which color is assigned to which level of the factor z

I don't know. I would take the approach of setting them manually, which has the added advantage of letting you chose your colour scheme. Add a call to scale_fill_manual and scale_colour_manual with a vector of colours.

I think the fill in your aesthetic (and consequently a call to scale_fill_manual) might be irrelevant as most shapes don't have a fill attribute, only a colour one. I'd check this but I don't have my PC to hand.

Alternatively, the default list of colours is available online at
http://www.cookbook-r.com/Graphs/Colors_(ggplot2)/
Although if you plan to do something like refer to a feature of your graph via something like 'the blue point in figure 1' you might find the default colours a challenge to describe/name. I do.

Regards,
Ron.

[Ron Crump]

Dear Ashim,

>Here is a reprex :

Thanks.

>x <- as.factor(rep(1:10,10))
>y <- rnorm(100)
>z <- as.factor(sample(x,100))
>
>ggplot(data = data.frame(x,y,z), aes(x=x,y=y,color = z,shape = z,fill =
>z))
>+ geom_point()
>

>I want to list out which color is assigned to which level of the factor z

I don't know. I would take the approach of setting them manually, which has the added advantage of letting you chose your colour scheme.

Alternatively, the default list of colours is probably available online

[Ron Crump]

My apologies for my most recent email. It was sent in error - I came across it in my drafts folder and thought I hadn't previously sent it.

Ron.

[Ashim Kapoor]

Dear Jianshuang and Ron,

Many thanks for your reply.

> library(scales)
> hue_pal()(10)
 [1] "#F8766D" "#D89000" "#A3A500" "#39B600" "#00BF7D" "#00BFC4" "#00B0F6"
 [8] "#9590FF" "#E76BF3" "#FF62BC"

> x <- as.factor(rep(1:10,10))
> y <- rnorm(100)
> z <- as.factor(sample(x,100))

> levels(z)
 [1] "1"  "2"  "3"  "4"  "5"  "6"  "7"  "8"  "9"  "10"
> 

Can I assume that the 10 colours generated by hue_pal will be mapped to the levels of z in the same order as they occur in levels(z)  ?

For instance if I do :

z <- factor(z, levels = seq(10,1))

> levels(z)
 [1] "10" "9"  "8"  "7"  "6"  "5"  "4"  "3"  "2"  "1"
>

In this case will the 10 colours reported by hue_pal()(10) be assigned to the 10 levels of z in the reverse of the original order?

Another query :

x <- as.factor(rep(1:10,10))
y <- rnorm(100)
z <- as.factor(sample(x,100))

gg1 <- ggplot(data = data.frame(x,y,z), aes(x=x,y=y,color = z,shape =z,fill = z)) + geom_point() +
       scale_shape_manual(values=c(11:20))

Note: we need the scale shape manual because we have more than 6 levels for shape. I forgot to do this in my orginal example (my apologies for this).

Can I not manipulate the following matrices and recover the mapping between colour and the levels of z?

> ggplot_build(gg1)$data[1]
[[1]]

    shape  colour    fill  x           y PANEL group size alpha stroke
1      16 #00BFC4 #00BFC4  1  0.29579631     1     4  1.5    NA    0.5
2      12 #D89000 #D89000  2 -0.43358638     1     8  1.5    NA    0.5
3      20 #FF62BC #FF62BC  3 -0.09725683     1    20  1.5    NA    0.5
4      15 #00BF7D #00BF7D  4  0.90603884     1    23  1.5    NA    0.5
5      17 #00B0F6 #00B0F6  5 -0.62904424     1    29  1.5    NA    0.5
6      18 #9590FF #9590FF  6 -1.07468368     1    37  1.5    NA    0.5
7      12 #D89000 #D89000  7 -1.23907191     1    40  1.5    NA    0.5
8      14 #39B600 #39B600  8 -0.40183759     1    49  1.5    NA    0.5
9      13 #A3A500 #A3A500  9  2.17610186     1    55  1.5    NA    0.5
10     16 #00BFC4 #00BFC4 10 -0.16761590     1    63  1.5    NA    0.5
11     20 #FF62BC #FF62BC  1 -0.55441511     1     7  1.5    NA    0.5
12     16 #00BFC4 #00BFC4  2 -0.91594616     1    12  1.5    NA    0.5
13     15 #00BF7D #00BF7D  3  0.30314967     1    17  1.5    NA    0.5
14     19 #E76BF3 #E76BF3  4  0.38348840     1    26  1.5    NA    0.5
15     14 #39B600 #39B600  5 -0.01784608     1    27  1.5    NA    0.5
16     15 #00BF7D #00BF7D  6 -0.82852709     1    35  1.5    NA    0.5
17     17 #00B0F6 #00B0F6  7  1.00782650     1    43  1.5    NA    0.5
18     12 #D89000 #D89000  8  0.45755324     1    47  1.5    NA    0.5
19     13 #A3A500 #A3A500  9 -0.02577137     1    55  1.5    NA    0.5
20     20 #FF62BC #FF62BC 10  0.85958394     1    65  1.5    NA    0.5
21     12 #D89000 #D89000  1 -0.06696429     1     2  1.5    NA    0.5

or the following matrix ?

> ggplot_build(gg1)$plot[[4]]
* x      -> x
* y      -> y
* colour -> z
* shape  -> z
* fill   -> z
> 

Which object contains the mapping between the levels of z and the colour assigned to it ?

Best Regards,

Ashim.