geom histogram labeling 25th,50th,75th percentiles

[help]

hello I have data that looks like this:

years n
1.11  19
1.33  3882
1.44. 321
1.91  1000
2.11  6844
2.33  2343 
3.44  348
...
...

I am seeking to make a histogram and label the 25th quantile, 50th(median), and 75th quantile with the numbers on the graph too. this is what i have so far:

clean_data <- data %>% uncount(weights = n)
ggplot(clean_data, aes(x = years)) +
geom_histogram() +
# this is code I found online to add 25th, 50th, and 75th percentiles with lines but the exact values are not present next to the lines
stat_summary(
aes(y = 1, xintercept = stat(x)),
fun.data = function(x) {
data.frame(
y = quantile(x, probs = c(0.25, 0.5, 0.75)) ) },
geom = "vline",
orientation = "y")

is there any better way to do this? thank you so much

[Flm]

@help

I am seeking to make a histogram and label the 25th quantile, 50th(median), and 75th quantile with the numbers on the graph too. this is what i have so far:

Could this be useful to you?

library(tidyverse)

data <- tribble(
  ~years, ~n,
  1.11,  19,
  1.33, 3882,
  1.44, 321,
  1.91, 1000,
  2.11, 6844,
  2.33, 2343,
  3.44, 348,
)

clean_data <- data %>% 
  uncount(weights = n)

q <- quantile(clean_data$years, probs = c(0.25, 0.5, 0.75))

Then you can make this plot

# plot 1
data %>% 
  ggplot(aes(x = years, y = n)) +
  geom_col() +
  geom_vline(xintercept = q[1]) +
  geom_vline(xintercept = q[2]) +
  geom_vline(xintercept = q[3]) +
  geom_label(aes(x = q[1], y = 7300, label = q[1])) +
  geom_label(aes(x = q[2], y = 7300, label = q[2])) +
  geom_label(aes(x = q[3], y = 7300, label = q[3])) +
  labs(x = "Years",
       y = "Count",
       title = "Title") +
  theme_classic()

Or something like this:

# plot 2
data %>% 
  ggplot(aes(x = years, y = n)) +
  geom_col() +
  geom_vline(xintercept = q[1]) +
  geom_vline(xintercept = q[2]) +
  geom_vline(xintercept = q[3]) +
  geom_label(aes(x = q[1], y = 7400, label = paste0("25%\n", q[1]))) +
  geom_label(aes(x = q[2], y = 7400, label = paste0("50%\n", q[2]))) +
  geom_label(aes(x = q[3], y = 7400, label = paste0("75%\n", q[3]))) +
  labs(x = "Years",
       y = "Count",
       title = "Title") +
  theme_classic()

[help]

thank u so much! is there a way to do this also with facet_wrap included so theres two different quantiles for each graph

[Flm]

which variable would you like to be divided by facet_wrap()?

[help]

@help

years n   animal
1.11  19      YES
1.33  3882   NO
1.44. 321     YES
1.91  1000   NO
2.11  6844   NO
2.33  2343    NO
1.11  19
1.33  3882
1.44. 321
1.91  1000
2.11  6844
2.33  2343 
3.44  348
3.44  348
..
...
.

if I had a binary variable called animal that was either yes or no, thank u so much

[Flm]

... +
facet_wrap(~animal, ncol = 1, strip.position = "r")

data %>% 
  ggplot(aes(x = years, y = n)) +
  geom_col() +
  geom_vline(xintercept = q[1]) +
  geom_vline(xintercept = q[2]) +
  geom_vline(xintercept = q[3]) +
  geom_label(aes(x = q[1], y = 7400, label = paste0("25%\n", q[1]))) +
  geom_label(aes(x = q[2], y = 7400, label = paste0("50%\n", q[2]))) +
  geom_label(aes(x = q[3], y = 7400, label = paste0("75%\n", q[3]))) +
  labs(x = "Years",
       y = "Count",
       title = "Title") +
  theme_classic() +
  facet_wrap(~animal, ncol = 1, strip.position = "r") +
  ylim(0, 8000)

[help]

thanks, but there will be a different percentile for the one graph vs another tho? not the same one exactly

[Axeman]

Perhaps slightly more idiomatic would be:

ggplot(...) +
  ... +
  geom_vline(xintercept = q) +
  annotate(geom = 'label', x = q, y = 7400, label = paste(c("25%", "50%", "75%"), q, sep = '\n'))

[help]

oh thank you axeman that makes sense

[help]

however I am still wondering how to get two different quantiles(25 percentile, 50th, 75) labeled on two different graphs with facet wrap. like each graph will have its own distribution

thank u all so much

[Flm]

maybe not the cleanest way, but it should work

library(tidyverse)
library(patchwork)

data <- tribble(
  ~years, ~n, ~animal,
  1.11,  19, "Y",
  1.33, 3882, "N",
  1.44, 321, "Y",
  1.91, 1000, "N",
  2.11, 6844, "N",
  2.33, 2343, "Y",
  3.44, 348, "Y"
)

clean_data <- data %>% 
  uncount(weights = n)


pl <- function(i){
  
  q <- clean_data %>% 
    filter(animal == i) %>%
    summarise(quantile(years, probs = c(0.25, 0.5, 0.75))) %>%
    pull()
  
  
  data %>% 
    filter(animal == i) %>%
    ggplot(aes(x = years, y = n)) +
    geom_col() +
    geom_vline(xintercept = q[1]) +
    geom_vline(xintercept = q[2]) +
    geom_vline(xintercept = q[3]) +
    geom_label(aes(x = q[1], y = max(n)+300, label = q[1])) +
    geom_label(aes(x = q[2], y = max(n)+300, label = q[2])) +
    geom_label(aes(x = q[3], y = max(n)+300, label = q[3])) +
    labs(x = "Years",
         y = "Count",
         title = paste("Type:", i)) +
    theme_classic() +
    theme(plot.title = element_text(hjust = .5, size = 18, face = "bold"))

}

pl("N") + pl("Y")

[help]

thank u so much! is this possible to do with geom_histogram too

[Flm]

Replica geom_col()

geom_histogram(stat = "identity")