glue function in R isn't printing the right value when used in ggplot2 title

Question

I'm writing a plot in ggplot2 using a ranking system and I want to plot the actual value in the title (as it's ranking appears in the plot), which later on I will use dlply to print multiple plots (one for each row of database). Here's an example of the plot, filtering to the tenth row (nregs == "1.10 Xingu e Interbacias Xingu - Tapajós"):

precipitation %>% 
        mutate(rank_precip = rank(-precip)) %>%
        filter(nregs == '1.10 Xingu e Interbacias Xingu - Tapajós') %>% 
        ggplot(aes(x = rank_precip, y = nregs)) +
        labs(title = glue("Accumulated Precipitation (1980-2010): {precip} mm")) +
        geom_segment(aes(x = 1, y = nregs, xend = 58, yend = nregs), size = 1.5, color = 'antiquewhite4') +
        geom_segment(data = expand.grid(x = rev(seq(1,58, by = 3)), y = 1),
                   aes(x = x, y = y, xend = x, yend = y + 0.05),
                   color = 'antiquewhite4',
                   size = 1.5) +
        geom_segment(data = expand.grid(x = rev(seq(1,58, by = 3)), y = 1),
                   aes(x = x, y = y, xend = x, yend = y - 0.05),
                   color = 'antiquewhite4',
                   size = 1.5) +
        scale_x_reverse() +
        geom_rect(aes(xmin = rank_precip - 0.5, xmax = rank_precip + 0.5, ymax = 1.1, ymin = 0.9), fill = 'blue3') +
        geom_rect(aes(xmin = 57.5, xmax = 58.5, ymax = 1.1, ymin = 0.9)) +
        geom_rect(aes(xmin = 0.5, xmax = 1.5, ymax = 1.1, ymin = 0.9)) +
        geom_text(aes(label = "58º", x = 57.75, y = 1.15), size = 6, fontface = 'bold') +
        geom_text(aes(label = "1º", x = 0.75, y = 1.15), size = 6, fontface = 'bold') +
        geom_rect(aes(xmin = rank_precip - 0.5, xmax = rank_precip + 0.5, ymax = 1.1, ymin = 0.9), fill = 'blue3') +
        geom_text(aes(x = rank_precip - 0.25, y = 1.15, label = glue("{rank_precip}º")), size = 6, fontface = 'bold', color = 'blue3') +
        theme_minimal() +
        theme(axis.text.y = element_blank(),
              axis.title.y = element_blank(),
              axis.text.x = element_blank(),
              axis.title.x = element_blank(),
              panel.grid = element_blank())

The plot printed is here:

from RStudio

However, as you can see from the dataset, the precipitation value in the title is printed as "67", but in fact it is 1937.08. I've tried to transform the dataset as tibble, transform the precip value into character but nothing worked, could you give me a help?

structure(list(nregs = c("1.1 Javari e Interbacias Javari - Juruá", 
"1.2 Transf. da Margem Esquerda do Solimões", "1.3 Juruá e Interbacias Juruá - Jutaí", 
"1.4 Purus e Interbacias Purus - Juruá", "1.5 Negro", "1.6 Madeira e Interbacias Madeira - Purus", 
"1.7 Estaduais Margem Esquerda do Amazonas", "1.8 Tapajós e Interbacias Tapajós - Madeira", 
"1.9 Estaduais PA", "1.10 Xingu e Interbacias Xingu - Tapajós", 
"1.11 Estaduais PA/AM", "1.12 Transfronteiriça Oiapoque", "1.13 Jari - Estaduais AP", 
"2.1 Araguaia", "2.2 Alto Tocantins", "2.3 Baixo Tocantins", 
"3.1 Gurupi - Estaduais MA", "3.2 Itapecuru", "4.1 Alto Parnaíba", 
"4.2 Médio Parnaíba", "4.3 Baixo Parnaíba", "5.1 Estaduais CE/PI", 
"5.2 Estaduais CE", "5.3 Estaduais CE/RN", "5.4 Piancó - Piranhas - Açu", 
"5.5 Estaduais RN", "5.6 Estaduais PB/PE", "5.7 Mundaú - Paraíba - Estaduais AL", 
"6.1 Alto São Francisco", "6.2 Médio São Francisco", "6.3 Sub Médio São Francisco", 
"6.4 Baixo São Francisco", "7.1 Vaza Barris - Real - Sergipe", 
"7.2 Itapecuru - Paraguaçu", "7.3 Contas", "7.4 Jequitinhonha - Pardo", 
"7.5 São Mateus - Mucuri - Itaúnas - Estaduais BA/MG", "8.1 Doce", 
"8.2 Estaduais ES", "8.3 Paraíba do Sul", "8.4 Estaduais RJ", 
"8.5 Estaduais SP", "8.6 Ribeira do Iguape", "9.1 Paranaíba", 
"9.2 Grande", "9.3 PCJ", "9.4 Tietê", "9.5 Paranapanema", "9.6 Paraná RH1", 
"9.7 Iguaçu", "10.1 Alto Paraguai", "10.2 Taquari - Miranda - Apa", 
"11.1 Alto Uruguai", "11.2 Médio Uruguai RS", "12.1 Estaduais PR", 
"12.2 Estaduais SC", "12.3 Guaíba e Estaduais RS", "12.4 Transfronteiriça Lagoa Mirim - Chuí"
), precip = c(2440.71, 2631.19, 2300.6, 2363.48, 2461.57, 2190.72, 
2297.28, 1960.7, 2095.36, 1937.08, 2400.93, 2387.93, 2420.15, 
1617.32, 1543.31, 2318.93, 2165.06, 1520.86, 1109.38, 872.52, 
1363.62, 1181.33, 844.2, 811.49, 826.64, 1197.12, 1094.41, 1261.29, 
1192.56, 883.14, 738.28, 823.37, 883.49, 876.4, 866.9, 950.23, 
1145.28, 1253.72, 1337.82, 1428.05, 1391.78, 1842.74, 1742.62, 
1511.54, 1517.9, 1549.69, 1459.11, 1549.54, 1503.42, 1764.78, 
1413.47, 1304.44, 1893.2, 1713.52, 2021.25, 1753.24, 1672.39, 
1412.73)), row.names = c(NA, -58L), class = c("tbl_df", "tbl", 
"data.frame"))

Jonny Phelps · Answer 1 · 2020-12-21T16:31:28.423

It's funny, it's using a base dataset called precip rather than your column. I'm not comfortable with glue so found an option with sprintf and used the answer from this SO post to format it ie the curly brackets and dot in ggplot call

library(tidyverse)
library(glue)

precipitation %>% 
  mutate(rank_precip = rank(-precip)) %>%
  filter(nregs == '1.10 Xingu e Interbacias Xingu - Tapajós') %>% 
  {
    ggplot(., aes(x = rank_precip, y = nregs)) +
      labs(title = sprintf("Accumulated Precipitation (1980-2010): %s mm", .$precip)) +
      geom_segment(aes(x = 1, y = nregs, xend = 58, yend = nregs), size = 1.5, color = 'antiquewhite4') +
      geom_segment(data = expand.grid(x = rev(seq(1,58, by = 3)), y = 1),
                   aes(x = x, y = y, xend = x, yend = y + 0.05),
                   color = 'antiquewhite4',
                   size = 1.5) +
      geom_segment(data = expand.grid(x = rev(seq(1,58, by = 3)), y = 1),
                   aes(x = x, y = y, xend = x, yend = y - 0.05),
                   color = 'antiquewhite4',
                   size = 1.5) +
      scale_x_reverse() +
      geom_rect(aes(xmin = rank_precip - 0.5, xmax = rank_precip + 0.5, ymax = 1.1, ymin = 0.9), fill = 'blue3') +
      geom_rect(aes(xmin = 57.5, xmax = 58.5, ymax = 1.1, ymin = 0.9)) +
      geom_rect(aes(xmin = 0.5, xmax = 1.5, ymax = 1.1, ymin = 0.9)) +
      geom_text(aes(label = "58º", x = 57.75, y = 1.15), size = 6, fontface = 'bold') +
      geom_text(aes(label = "1º", x = 0.75, y = 1.15), size = 6, fontface = 'bold') +
      geom_rect(aes(xmin = rank_precip - 0.5, xmax = rank_precip + 0.5, ymax = 1.1, ymin = 0.9), fill = 'blue3') +
      geom_text(aes(x = rank_precip - 0.25, y = 1.15, label = glue("{rank_precip}º")), size = 6, fontface = 'bold', color = 'blue3') +
      theme_minimal() +
      theme(axis.text.y = element_blank(),
            axis.title.y = element_blank(),
            axis.text.x = element_blank(),
            axis.title.x = element_blank(),
            panel.grid = element_blank()) 
  }

EDIT:

as a function to do for each level of nregs:

rank_precip_plot_fun <- function(input){
  ggplot(input, aes(x = rank_precip, y = nregs)) +
    labs(title = sprintf("Accumulated Precipitation (1980-2010): %s mm", input$precip)) +
    geom_segment(aes(x = 1, y = nregs, xend = 58, yend = nregs), size = 1.5, color = 'antiquewhite4') +
    geom_segment(data = expand.grid(x = rev(seq(1,58, by = 3)), y = 1),
                 aes(x = x, y = y, xend = x, yend = y + 0.05),
                 color = 'antiquewhite4',
                 size = 1.5) +
    geom_segment(data = expand.grid(x = rev(seq(1,58, by = 3)), y = 1),
                 aes(x = x, y = y, xend = x, yend = y - 0.05),
                 color = 'antiquewhite4',
                 size = 1.5) +
    scale_x_reverse() +
    geom_rect(aes(xmin = rank_precip - 0.5, xmax = rank_precip + 0.5, ymax = 1.1, ymin = 0.9), fill = 'blue3') +
    geom_rect(aes(xmin = 57.5, xmax = 58.5, ymax = 1.1, ymin = 0.9)) +
    geom_rect(aes(xmin = 0.5, xmax = 1.5, ymax = 1.1, ymin = 0.9)) +
    geom_text(aes(label = "58º", x = 57.75, y = 1.15), size = 6, fontface = 'bold') +
    geom_text(aes(label = "1º", x = 0.75, y = 1.15), size = 6, fontface = 'bold') +
    geom_rect(aes(xmin = rank_precip - 0.5, xmax = rank_precip + 0.5, ymax = 1.1, ymin = 0.9), fill = 'blue3') +
    geom_text(aes(x = rank_precip - 0.25, y = 1.15, label = glue("{rank_precip}º")), size = 6, fontface = 'bold', color = 'blue3') +
    theme_minimal() +
    theme(axis.text.y = element_blank(),
          axis.title.y = element_blank(),
          axis.text.x = element_blank(),
          axis.title.x = element_blank(),
          panel.grid = element_blank()) 
}


dt <- precipitation %>% 
  mutate(rank_precip = rank(-precip))

library(plyr)
plots <- dlply(dt, .(nregs), rank_precip_plot_fun)
plots[[1]]
plots[[2]] # different value

it does partially the job! but when i remove the filter and save the ggplot code as a function "p", and apply this code: ```require(plyr) plots = dlply(precipitation, .(nregs), function (x) p %+% x) invisible(mapply(ggsave, file = paste0(names(plots),".jpeg"), plot = plots)) ``` it actually just keeps printing the precip value from the 1st row of the dataset (2440.71) — fernando schuh, Dec 21 '20 at 15:52
I've edited it with a functional option. The ranking has to happen before you use `dlply` and so the only command that's left is the `ggplot2` so the `.` is no longer needed and can just pass in data as usual — Jonny Phelps, Dec 21 '20 at 16:32

glue function in R isn't printing the right value when used in ggplot2 title

1 Answers1