为多行分配颜色并在ggplot中添加图例

Question

I successfully created multiple line plots on a single plot using ggplot. However, I am having difficulty adjusting the colors for each line, and then creating/formatting a legend corresponding to these colors. First, here is the data that I am working with (contained in four separate data frames):

newdata<-data.frame(RCP1pctCO2cumulative, Column) #length 140
newdata1<-data.frame(RCP8.5cumulative, Column1) #length 90
newdata2<-data.frame(RCP4.5cumulative, Column2) #length 90
newdata3<-data.frame(Historicalcumulative, Column3) #length 145

head(newdata) 

RCP1pctCO2cumulative   Column
layer.1          0.000000000 30.62975
layer.2          0.006974906 30.29012
layer.3          0.013907599 30.43212
layer.4          0.021697436 30.70810
layer.5          0.030232970 30.38155
layer.6          0.038998084 30.34130

head(newdata1)

RCP8.5cumulative  Column1
layer.1        0.4475691 31.94422
layer.2        0.4569296 31.93002
layer.3        0.4663113 31.98923
layer.4        0.4756628 32.16458
layer.5        0.4850761 32.20246
layer.6        0.4946258 32.16779

head(newdata2)

RCP4.5cumulative  Column2
layer.1        0.4487829 32.05137
layer.2        0.4584334 32.01951
layer.3        0.4680946 32.04347
layer.4        0.4777492 32.23928
layer.5        0.4875477 32.61044
layer.6        0.4974490 32.14446

head(newdata3)

Historicalcumulative  Column3
layer.1         0.0000000000 30.20725
layer.2         0.0009499752 30.30651
layer.3         0.0017818766 30.35118
layer.4         0.0025179833 30.13334
layer.5         0.0031186696 30.14842
layer.6         0.0036720898 29.87103

I tried the following:

gg<-ggplot(newdata, aes(x=RCP1pctCO2cumulative, y=Column)) + 
geom_smooth(), color="black")

gg + geom_smooth(data=newdata1, aes(x=RCP4.5cumulative, y=Column1)) + 
geom_smooth(data=newdata2, aes(x=RCP8.5cumulative, y=Column2)) + 
geom_smooth(data=newdata3, aes(x=Historicalcumulative, y=Column3)), 
color="red", size=3)

This yields this error:

Error: unexpected ',' in "gg + geom_smooth(data=newdata1, 
aes(x=RCP4.5cumulative, y=Column1)) + geom_smooth(data=newdata2, 
aes(x=RCP8.5cumulative, y=Column2)) + geom_smooth(data=newdata3, 
aes(x=Historicalcumulative, y="

I know that there is a comma placed incorrectly somewhere, but I suspect that will not correct the overall problem of trying to assign specific colors to each line plot. Ideally, for "gg", I would like that line to be "green", while for the second, third and fourth lines, I want those to be "blue", "red" and "black", respectively. Then, I want to show and be able to format this in a legend, showing what each color represents.

Any help with this would be extremely valuable! It seems straight-forward, but I think the problem is related to where the "color" command needs to be placed for each line plot, and then constructing a legend.

Thanks,

Answer 1

Starting with the comma issue (and then let me know if the rest works out!)

gg<-ggplot(newdata, aes(x=RCP1pctCO2cumulative, y=Column)) + 
geom_smooth(color="#000000")

gg + geom_smooth(data=newdata1, aes(x=RCP4.5cumulative, y=Column1)) + 
geom_smooth(data=newdata2, aes(x=RCP8.5cumulative, y=Column2)) + 
geom_smooth(data=newdata3, aes(x=Historicalcumulative, y=Column3), 
color="red", size=3)

In case I misunderstood your data, here is an example using the famous Iris data:

ggplot(data=iris, aes(x=Sepal.Length, y=Sepal.Width)) + 
geom_smooth(color="#000000") + 
geom_smooth(data=iris, aes(x=Petal.Length, y=Petal.Width), color="#FF0000", size=3)

To get a legend, by far the easiest way is presented in this answer . This gist of it is, use melt from reshape2 to get all values into one dataframe with a categorical variable that determines color.

Answer 2

Edit: using OP data

Getting a legend set up in ggplot for data from multiple tables will be much easier if you can combine them into one table with common variables, where the source table is noted in another variable.

library(tidyverse)
# Convenience function to rename columns to something uniform
make_uniform <- function(df, sourcename) {
  df %>%
    rename(x = 2, y = 3) %>% # Rename 2nd and 3rd columns
    mutate(source = sourcename)
}

combined <- bind_rows(
  RCP1pctCO2cumulative %>% make_uniform("CO2"),
  RCP8.5cumulative     %>% make_uniform("RCP8.5"), 
  RCP4.5cumulative     %>% make_uniform("RCP4.5"), 
)

> combined
   rowname           x        y source
1  layer.1 0.000000000 30.62975    CO2
2  layer.2 0.006974906 30.29012    CO2
3  layer.3 0.013907599 30.43212    CO2
4  layer.4 0.021697436 30.70810    CO2
5  layer.5 0.030232970 30.38155    CO2
6  layer.6 0.038998084 30.34130    CO2
7  layer.1 0.447569100 31.94422 RCP8.5
8  layer.2 0.456929600 31.93002 RCP8.5
9  layer.3 0.466311300 31.98923 RCP8.5
10 layer.4 0.475662800 32.16458 RCP8.5
11 layer.5 0.485076100 32.20246 RCP8.5
12 layer.6 0.494625800 32.16779 RCP8.5
13 layer.1 0.448782900 32.05137 RCP4.5
14 layer.2 0.458433400 32.01951 RCP4.5
15 layer.3 0.468094600 32.04347 RCP4.5
16 layer.4 0.477749200 32.23928 RCP4.5
17 layer.5 0.487547700 32.61044 RCP4.5
18 layer.6 0.497449000 32.14446 RCP4.5

ggplot(combined, aes(x, y, color = source)) +
  geom_smooth()

---

loading data from OP

RCP1pctCO2cumulative  <- read.table(
  header = T, 
  stringsAsFactors = F,
  text = "rowname RCP1pctCO2cumulative   Column
layer.1          0.000000000 30.62975
layer.2          0.006974906 30.29012
layer.3          0.013907599 30.43212
layer.4          0.021697436 30.70810
layer.5          0.030232970 30.38155
layer.6          0.038998084 30.34130")


RCP8.5cumulative  <- read.table(
  header = T, 
  stringsAsFactors = F,
  text = "rowname RCP8.5cumulative  Column1
layer.1        0.4475691 31.94422
layer.2        0.4569296 31.93002
layer.3        0.4663113 31.98923
layer.4        0.4756628 32.16458
layer.5        0.4850761 32.20246
layer.6        0.4946258 32.16779")



RCP4.5cumulative    <- read.table(
  header = T, 
  stringsAsFactors = F,
  text = "rowname RCP4.5cumulative  Column2
layer.1        0.4487829 32.05137
layer.2        0.4584334 32.01951
layer.3        0.4680946 32.04347
layer.4        0.4777492 32.23928
layer.5        0.4875477 32.61044
layer.6        0.4974490 32.14446")