geom_rect: background color repeated per season

Question

I have a dataframe like this:

df<-data.frame(Category= c("a","b","a","b"), Value = c(25,90,40,10), Date= c("2016-02-13", "2016-05-13", "2016-08-13", "2016-11-13"))

In reality it is more complex, has several years and several observed objects so that it should be a faceted plot in the end, but I think this has nothing to do with the question.

I want to have a ggplot (line plot), where every season got it's own background color. e.g.: spring from March to May in yellow, summer from June to August in red autumn from September to November in blue and winter from December to February in grey.

This should be repeated, regardless the year as it goes through several years and the database will be updated with time.

I tried a lot with geom_rect but didn't find a working solution.

Thanks for any advice!

Answer 1

If I understand your goal correctly, I think you can achieve it by creating two additional variables, say Season and Color that correspond to Date column, and then supply the columns as necessary to geom_line.

To make the steps and the results clearer, I create a dummy data by expanding your data to another year (2017) with similar date and category but slightly different values:

1. The data

df<-data.frame(Category= c("a","b","a","b"), Value = c(25,90,40,10), Date= c("2016-02-13", "2016-05-13", "2016-08-13", "2016-11-13"))
df2<-data.frame(Category= c("a","b","a","b"), Value = c(30,95,45,15), Date= c("2017-02-13", "2017-05-13", "2017-08-13", "2017-11-13"))
dat <- rbind.data.frame(df,df2)

dat
  Category Value       Date
1        a    25 2016-02-13
2        b    90 2016-05-13
3        a    40 2016-08-13
4        b    10 2016-11-13
5        a    30 2017-02-13
6        b    95 2017-05-13
7        a    45 2017-08-13
8        b    15 2017-11-13

2. Creating Season and Color columns

dat.season <- dat %>% 
     mutate(Date = as.Date(Date)) %>% 
     mutate(Month = months(Date)) %>% 
     mutate(Season = case_when(Month %in% c("March", "April", "May") ~ "spring",  
                               Month %in% c("June", "July", "August") ~ "summer", 
                               Month %in% c("September", "October", "November") ~ "autumn", 
                               Month %in% c("December", "January", "February")~ "winter")) %>% 
     mutate(Color = case_when(Season == "spring"~ "yellow", 
                              Season == "summer"~ "red", 
                              Season == "autumn"~ "blue", 
                              Season == "winter"~ "grey"))
dat.season
  Category Value       Date    Month Season  Color
1        a    25 2016-02-13 February winter   grey
2        b    90 2016-05-13      May spring yellow
3        a    40 2016-08-13   August summer    red
4        b    10 2016-11-13 November autumn   blue
5        a    30 2017-02-13 February winter   grey
6        b    95 2017-05-13      May spring yellow
7        a    45 2017-08-13   August summer    red
8        b    15 2017-11-13 November autumn   blue

Supplying the columns to geom_line()

dat.season %>% ggplot() + 
  geom_line(aes(x = Date, y = Value), 
            colour = dat.season$Color) + 
  theme_bw()

The result

Update to add coloured background

Here is the line plot along with coloured backgrounds for each season.

dat.season %>% 
    ggplot() + 
    geom_rect(aes(xmin = Date[1], xmax = Date[2], 
                  ymin = Value[1], ymax = Value[2]), 
              fill = dat.season$Color[1])+
    geom_rect(aes(xmin = Date[2], xmax = Date[3], 
                  ymin = Value[2], ymax = Value[3]), 
              fill = dat.season$Color[2])+
    geom_rect(aes(xmin = Date[3], xmax = Date[4], 
                  ymin = Value[3], ymax = Value[4]), 
              fill = dat.season$Color[3])+
    geom_rect(aes(xmin = Date[4], xmax = Date[5], 
                  ymin = Value[4], ymax = Value[5]), 
              fill = dat.season$Color[4])+
    geom_rect(aes(xmin = Date[5], xmax = Date[6], 
                  ymin = Value[5], ymax = Value[6]), 
              fill = dat.season$Color[5])+
    geom_rect(aes(xmin = Date[6], xmax = Date[7], 
                  ymin = Value[6], ymax = Value[7]), 
              fill = dat.season$Color[6])+
    geom_rect(aes(xmin = Date[7], xmax = Date[8], 
                  ymin = Value[7], ymax = Value[8]), 
              fill = dat.season$Color[7])+
    geom_line(aes(x = Date, y = Value)) +
    theme_bw()

The result