r - Create a separate ggplot legend for a ribbon behind the main data - Stack Overflow

admin管理员组

文章数量:1344238

I have a chart showing the observed values, a prediction for each value, and a ribbon behind it indicating the error region of the prediction. Something along these lines:

library(dplyr)
library(lubridate)
library(tidyr)
library(ggplot2)

set.seed(1234)

df <- economics %>%
  mutate(year = year(date)) %>%
  group_by(year) %>%
  summarise(pop = mean(pop, na.rm = TRUE)) %>%
  ungroup() %>%
  select(year, pop) %>%
  rowwise() %>%
  mutate(pred_factor = runif(1, 0, 2)) %>%
  mutate(limit_factor = runif(1, 0.5, 0.8)) %>%
  ungroup() %>%
  mutate(pred = pop * pred_factor) %>%
  mutate(limit_distance = pred * limit_factor) %>%
  mutate(lower_limit = pred - limit_distance) %>%
  mutate(upper_limit = pred + limit_distance) %>%
  select(year, pop, pred, lower_limit, upper_limit)

df %>%
  pivot_longer(cols = c(pop, pred), names_to = 'metric', values_to = 'value') %>%
  ggplot(aes(x = year, y = value, 
             ymin = lower_limit, ymax = upper_limit, 
             linetype = metric)) +
  geom_ribbon(fill = 'lightblue') +
  geom_line() +
  theme_bw()

However, I would like to separate out the ribbon colour into its own legend item, so I can label it separately. The best I have come up with so far is this:

df %>%
  pivot_longer(cols = c(pop, pred), names_to = 'metric', values_to = 'value') %>%
  mutate(shading = 'predicted range') %>%
  ggplot(aes(x = year, y = value, 
             ymin = lower_limit, ymax = upper_limit, 
             linetype = metric, fill = shading)) +
  geom_ribbon() +
  geom_line() +
  scale_fill_manual(values = c('predicted range' = 'lightblue')) +
  theme_bw()

But for some reason this approach creates a dark grey fill behind the two 'metric' items in this one, rendering them essentially illegible. Is there a different way I can arrange this data and chart to have this functional?

I have a chart showing the observed values, a prediction for each value, and a ribbon behind it indicating the error region of the prediction. Something along these lines:

library(dplyr)
library(lubridate)
library(tidyr)
library(ggplot2)

set.seed(1234)

df <- economics %>%
  mutate(year = year(date)) %>%
  group_by(year) %>%
  summarise(pop = mean(pop, na.rm = TRUE)) %>%
  ungroup() %>%
  select(year, pop) %>%
  rowwise() %>%
  mutate(pred_factor = runif(1, 0, 2)) %>%
  mutate(limit_factor = runif(1, 0.5, 0.8)) %>%
  ungroup() %>%
  mutate(pred = pop * pred_factor) %>%
  mutate(limit_distance = pred * limit_factor) %>%
  mutate(lower_limit = pred - limit_distance) %>%
  mutate(upper_limit = pred + limit_distance) %>%
  select(year, pop, pred, lower_limit, upper_limit)

df %>%
  pivot_longer(cols = c(pop, pred), names_to = 'metric', values_to = 'value') %>%
  ggplot(aes(x = year, y = value, 
             ymin = lower_limit, ymax = upper_limit, 
             linetype = metric)) +
  geom_ribbon(fill = 'lightblue') +
  geom_line() +
  theme_bw()

However, I would like to separate out the ribbon colour into its own legend item, so I can label it separately. The best I have come up with so far is this:

df %>%
  pivot_longer(cols = c(pop, pred), names_to = 'metric', values_to = 'value') %>%
  mutate(shading = 'predicted range') %>%
  ggplot(aes(x = year, y = value, 
             ymin = lower_limit, ymax = upper_limit, 
             linetype = metric, fill = shading)) +
  geom_ribbon() +
  geom_line() +
  scale_fill_manual(values = c('predicted range' = 'lightblue')) +
  theme_bw()

But for some reason this approach creates a dark grey fill behind the two 'metric' items in this one, rendering them essentially illegible. Is there a different way I can arrange this data and chart to have this functional?

• r
• ggplot2

Share Improve this question Follow asked 3 hours ago MargaretMargaret 5,9392222 gold badges5858 silver badges7676 bronze badges

Add a comment  | 

1 Answer 1

Sorted by: Reset to default 1

Just move the ribbon and line-specific aesthetics out of ggplot() and into their respective geoms:

library(dplyr)
library(tidyr)
library(ggplot2)

df %>%
  pivot_longer(cols = c(pop, pred), names_to = 'metric', values_to = 'value') %>%
  mutate(shading = 'predicted range') %>%
  ggplot(aes(x = year, y = value)) +
  geom_ribbon(aes(ymin = lower_limit, ymax = upper_limit, fill = shading)) +
  geom_line(aes(linetype = metric)) +
  scale_fill_manual(values = c('predicted range' = 'lightblue')) +
  theme_bw()

(You’ll actually get the result you want just by moving aes(linetype = metric) into geom_line(), but I moved the geom_ribbon() aesthetics down as well for consistency.)

本文标签： rCreate a separate ggplot legend for a ribbon behind the main dataStack Overflow