How to add months on top of daily data on the x-axis in ggplot2?

Question

I have a dataframe that has precipitation (cm) for Y and days for X.

> head(precip)
  Day     mean
1   1 1.784654
2   2 1.850796
3   3 1.850486
4   4 1.858424
5   5 1.884435
6   6 1.907781

until 366 days (averages include leap years). Both columns are currently as numeric. This is my code for line plotting it:

precip_plots <- ggplot(data = precip, aes_string(x = "Day", y = "mean", group=1)) +
  geom_line(size = 1)+
  theme_bw()+
  labs(title = "Daily precipitation",
      x = "Day",
      y = "Precipitation (cm)")

Which looks like

I want the days to be on the X-axis, but below (also on X-axis) I want there to be month letters (JFMAMJJASOND) for each range of the month, like

I have tried adding this to the ggplot:

scale_x_discrete(limits = 1:366, labels=Day[1:31] = "J", Day[32:60] = "F", Day[61:91] = "M", 
                   Day[92:121] = "A", Day[122:152] = "M", Day[153:182] = "J", Day[183:213] = "J",
                   Day[214:244] = "A", Day[245:274] = "S", Day[275:305] = "O", Day[306:335] = "N", 
                   Day[336:366] = "D")

but it just returns a bunch of "Error: unexpected '='" and "Error: unexpected ','".

I've looked everywhere online and on SO and I can't seem to find anyone with a similar issue.

Answer 1

Update
Here is one way to do it using a function as labels argument in scale_x_continuous. This works only if the breaks contains all days as well as the regular breaks (100, 200 etc.). Since this messes up the break lines and axis ticks (they are not symmetrically anymore) I've hidden them with element_blank() and inserted custom break lines with geom_vline.

library(tidyverse)

# some toy data similar to the original data
dat <- tibble(Day = 1:366,
              mean = 2+rnorm(366))

# setup breaks
month <- c(31,60,91,121,152,182,213,244,274,305,335,365)
breaks <- seq(100,300,100)
mybreaks <- sort(c(month,breaks))

precip_plots <- ggplot(data = dat,
                       aes_string(x = "Day",
                                  y = "mean",
                                  group = 1)) +
  geom_line(size = 1) +
  scale_x_continuous(
    breaks = mybreaks,
    # custom function using `dplyr::case_when`
    labels = function(x) {
      
      case_when(
        x == 31  ~ "\nJ",
        x == 60  ~ "\nF",
        x == 91 ~ "\nM",
        x == 121 ~ "\nA",
        x == 152 ~ "\nM",
        x == 182 ~ "\nJ",
        x == 213 ~ "\nJ",
        x == 244 ~ "\nA",
        x == 274 ~ "\nS",
        x == 305 ~ "\nO",
        x == 335 ~ "\nN",
        x == 365 ~ "\nD",
        x %% 100 == 0 ~ as.character(x))
      
    }) +
  # creates fake break lines
  geom_vline(xintercept = breaks,
             color = "#d6d6d6",
             size = 0.5) + 
  theme_bw() +
  # hides original break lines and axis ticks
  theme(panel.grid.major.x = element_blank(),
        panel.grid.minor.x = element_blank(),
        axis.ticks.x = element_blank()
  ) +
  labs(title = "Daily precipitation",
       x     = "Day",
       y     = "Precipitation (cm)") 

precip_plots

^{Created on 2021-08-07 by the reprex package (v0.3.0)}