My plot has three layers and won't let me change the legend?

Question

So I'm trying to change the legend on the right side to correspond to a specific date. I wanna change the Legend title to "Dates" and the color red to "0724", green to "0802", and blue to "0815". I've tried using scale_fill_discrete but it doesn't seem to work. I'm not sure if importing three data frames into this plot has something to do with it.

This is the code I'm currently using

ggplot(data = dataS24, aes(x = NominalKM, y = MassDefectrounded)) +
  geom_point(shape = 16, alpha = 0.7, aes(color = "#00A5FF")) +
  geom_point(data = dataS02, shape = 16, alpha = 0.7, aes(color = "#00BC59"))+
  geom_point(data = dataS15, shape = 16, alpha = 0.7, aes(color = "#FC717F"))+
guides(shape = guide_legend(order = 1))+
theme(panel.grid.major = element_blank(), 
                  panel.grid.minor = element_blank(),
                  panel.background = element_blank(),
                  axis.line = element_line(colour = "black"),
                  axis.text = element_text(size = 11, face = "bold"),
                  axis.title = element_text(size=13, face="bold"),
                  aspect.ratio = 1,
      )

I'm not sure where to integrate the code to change the legend. Thanks in advance!

This is pretty much how all three data frames look like, but they have at least 1000 rows and more columns with unnecessary information.

#data0724
Molecular.Formula  NominalKM  MassDefectRounded
C19H14O6S2         402        0.42889912
C19H14O8S2         433        0.47350283

Have a read of this: http://www.cookbook-r.com/Graphs/Colors_(ggplot2)/ — tjebo, Feb 13 '20 at 23:35

dc37 · Answer 1 · 2020-02-14T00:25:23.807

Without a reproducible example, it is hard to be able to assist you. A better practice will be to merge your three dataframe and create a categorical variable to separate them.

However, if you are looking to change the title and labels of the legend, you can use scale_color_manual (you will have to use scale_fill_manual only if you have pass your argument using fill = in the aes, here as you use color, it is scale_color_manual needed):

ggplot(data = dataS24, aes(x = NominalKM, y = MassDefectrounded)) +
  geom_point(shape = 16, alpha = 0.7, aes(color = "#00A5FF")) +
  geom_point(data = dataS02, shape = 16, alpha = 0.7, aes(color = "#00BC59"))+
  geom_point(data = dataS15, shape = 16, alpha = 0.7, aes(color = "#FC717F"))+
  guides(shape = guide_legend(order = 1))+
  theme(panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        axis.line = element_line(colour = "black"),
        axis.text = element_text(size = 11, face = "bold"),
        axis.title = element_text(size=13, face="bold"),
        aspect.ratio = 1,
  )+
  scale_color_manual(name = "Date", values = c("#00A5FF" = "0724", "#00BC59" = "0802", "#FC717F" = "0815"))

If this is not working, please consider providing a small reproducible example of your three dataframes (read this link: How to make a great R reproducible example)

NB: I'm not sure that this part is really essential to your plot: guides(shape = guide_legend(order = 1))+... as all shapes equal to 16 and are not set into an aes. But, I'm maybe wrong (I can't verify it without an example of your data).

EDIT: Merging three dataframes

Assuming that all your dataframes have the same format as the one you provided as example, here I generate three random dataframe that mimicks the structure of yours:

data0724 = data.frame(Molecular = LETTERS[1:5],
                      Nominal = sample(100:750,5),
                      Mass = rnorm(5, mean = 1, sd = 1)) 
data0802 = data.frame(Molecular = LETTERS[6:10],
                    Nominal = sample(100:750,5),
                    Mass = rnorm(5, mean = 1, sd = 1)) 
data0815 = data.frame(Molecular = LETTERS[11:15],
                      Nominal = sample(100:750,5),
                      Mass = rnorm(5, mean = 1, sd = 1))

Then, to merge them, I will add a categorical column that will display their name, then I used rbind to associate them:

data0724$Grp = "0724"
data0802$Grp = "0802"
data0815$Grp = "0815"

DF <- rbind(data0724,data0802,data0815)

   Molecular Nominal       Mass  Grp
1          A     623  0.8796062 0724
2          B     739  2.0128343 0724
3          C     531  0.7985419 0724
4          D     381 -1.0376825 0724
5          E     303  0.8041108 0724
6          F     244  2.2765787 0802
7          G     310  0.7750387 0802
8          H     168  0.6781074 0802
9          I     448  2.4878378 0802
10         J     576 -0.6679280 0802
11         K     296  1.2796279 0815
12         L     493  2.8778640 0815
13         M     391  0.9959393 0815
14         N     491  0.7215460 0815
15         O     354  1.4749117 0815

So, now if you want to plot them, you don't have to use a line for eahc dataframe but rather do:

library(ggplot2)
ggplot(DF, aes(x = Nominal, y = Mass, color = Grp))+
  geom_point(shape = 16, alpha = 0.7)+
  theme(panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        panel.background = element_blank(),
        axis.line = element_line(colour = "black"),
        axis.text = element_text(size = 11, face = "bold"),
        axis.title = element_text(size=13, face="bold"),
        aspect.ratio = 1,
  )+
  scale_color_manual(name = "Date", values = c("#00A5FF","#00BC59","#FC717F"))

And you get:

Does it answer your question ?

Edited above is an example of my data frame. – David Feb 14 '20 at 00:04 — David, Feb 14 '20 at 00:04

My plot has three layers and won't let me change the legend?

1 Answers1