Ploting red deciles in Q-Q plot stat_qq() or geom

I'm trying to compare a distribution against theoretical normal distribution but I would like to represents the decimal quantiles as red in the Q-Q plot.

This is the numeric array I'm using (as a df).

structure(list(v1 = c(84.01, 86.88, 48.5, 80.28, 63.85, 81.95, 
8.94, 59.16, 35.16, 49.6, 56.16, 19.71, 63.03, 64.82, 2.92, 46.69, 
42, 63.83, 53.29, 67.79, 15.87, 37.95, 52.99, 52.11, 66.14, 68.61, 
61.19, 44.9, 54.8, 61.81, 80.9, 26.17, 12.7, 57.51, 73.41, 54.25, 
55.61, 55.64, 57.69, 67.77, 23.21, 74.23, 29.43, 52.08, 41.1, 
58.65, 55.25, 79.08, 71.39, 32.29, 66.9, 60.52, 100, 43.34, 47.69, 
65.64, 75.9, 59.95, 55.12, 37.57, 60.76, 79.82, 55.67, 43.17, 
56.56, 39.87, 65.32, 53.82, 57.76, 26.88, 62.83, 42.71, 73.82, 
55.02, 73.52, 68.9, 60.82, 52.16, 61.34, 57.59, 19.71, 74.17, 
63.93, 59.88, 55.47, 43.19, 45.17, 64.23, 99.18, 53.25, 58, 73.66, 
100, 58.83, 86.6, 54.83, 58.13, 68.57, 56.17, 59.73, 47.02, 33.97, 
60.2, 60.53, 47.18, 54.72, 49.71, 35.29, 62.25, 44.07, 46.35, 
17.29, 26.93, 60.26, 77.35, 59.18, 71.83, 72.01, 58.78, 29.14, 
62.56, 23.41, 39.82, 56.27, 46.55, 60.05, 80.74, 15.99, 58.64, 
37.43, 60.38, 48.77, 54.41, 49.4, 36.05, 35.17, 80.04, 31.64, 
56.2, 39.35, 10.61, 75.99, 51.69, 59.65, 52.58, 64.76, 76.07, 
43.96, 66.95, 29.45, 43.18, 30.99, 51.22, 30.45, 73.49, 58.92, 
67.22, 51.68, 62.56, 64.29, 46.11, 72.86, 66.33, 56.27, 62.04, 
49.68, 46.44, 49.3, 66.93, 43.86, 12.72, 44.43, 60.62, 33.98, 
83.57, 50.37, 36.96, 54.04, 30.47, 28.46, 48.35, 35.64, 71.47, 
76.1, 7.32, 3.36, 83.27, 52.11, 90.43, 22.85, 7.9, 52.19, 41.96, 
23.22, 42.86, 30.83, 54.66, 63.9, 46.04, 43.69, 84.14, 50.09, 
65.29, 42.9, 72.92, 60.2, 45.36, 51.76, 58.7, 36.84, 77.72, 21.04, 
53.21, 68.87, 12.22, 63.93, 18.57, 74.93, 65.3, 79.18, 38.03, 
55.69, 73.93, 63.33, 93.97, 34.99, 45.87, 61.12, 47.95, 40.82, 
54.1, 69.17, 54.9, 41.72, 76.39, 61.75, 79.3, 29.62, 41.29, 33.33, 
52.31, 31.04, 65.94, 23.29, 64.27, 63.71, 73.43, 68.94, 68.85, 
68.39, 39.13, 65.16, 61.77, 88.54, 76.29, 78.85, 44.97, 12.35, 
90.31, 63.24, 77.85, 58.5, 81.05, 85.38, 74.16, 78.56, 54.86, 
71.15, 62.7, 64.49, 27.32, 45.71, 56.29, 60.76, 90, 63.91, 71.72, 
20.42, 45.02, 14.21, 54.64, 68.88, 71.67, 37.67, 53.88, 58.8, 
33.86, 57.98, 57.84, 73.74, 39.83, 40.69, 54.66, 18.08, 43.79, 
31.05, 89.34, 46.65, 24.99, 64.33, 51.87, 65.17, 40.04, 69.41, 
13.89, 61.99, 44.14, 65.46, 47.14, 27.88, 52.46, 12.69, 45.41, 
84.13, 59.48, 40.2, 40.76, 42.66, 35.56, 57.52, 41.17, 60.02, 
44.58, 1.76, 35.8, 70.61, 20.74, 1.78, 60.11, 70.64, 26.04, 33.58, 
61.88, 12.38, 48.17, 27.09, 75.23, 28.85, 53.25, 44.19, 64.56, 
17.12, 51.22, 66.02, 43.06, 58.71, 55.65, 27.52, 45.28, 44.83, 
74.25, 58.35, 23.1, 58.52, 54.1, 57.74, 47.61, 17.25, 57.44, 
70.38, 43.69, 89.84, 72.9, 57.53, 33.53, 18.08, 41.71, 52.07, 
46.63, 42.64, 41.98, 27.74, 72.13, 61.92, 83.65, 22.42, 62.47, 
13.49, 68.39, 48.94, 72.79, 49.76, 63.05, 85, 45.58, 59.58, 57.94, 
65.93, 65.26, 35.56, 65.71, 62.17, 21.07, 53.33, 72.97, 36.32, 
55.38, 68.96, 66.11, 56.05, 70.68, 89.05, 73.7, 85.43, 88.74, 
51.16, 49.83, 63.85, 45.61, 29.15, 49.93, 62.9, 61.39, 69.17, 
6.61, 63.66, 76.21, 56.85, 49.86, 68.41, 75.94, 67.88, 62.25, 
26.8, 74.27, 76.48, 46.25, 73.84, 47.21, 60.76, 42.17, 45.73, 
41.4, 67.04, 48.28, 63.73, 65.9, 36.62, 74.91, 50.11, 24.21, 
91.45, 78.71, 100, 27.59, 78.42, 38.69, 21.68, 0, 44.85, 59.28, 
45.75, 36.84, 73.91, 41.8, 66, 74.88, 59.25, 54.7, 44.39, 40.95, 
42.03, 66.93, 38, 72.08, 48, 41.7, 72.1, 64.23, 14.18, 43.53, 
57.81, 52.68, 45.73, 97.12, 18.98, 57.12, 48.33, 18.94, 60.6, 
52.05, 67.09, 72.89, 64.16, 57.79, 45.62, 55.02, 48.86, 60.74, 
7.35, 56.31, 60.51, 49, 67.56, 84.62, 15.98, 54.67, 65.72, 65.79, 
70.55, 74.18, 55.72, 57.89, 54.67, 39.83, 21.2, 78)), class = "data.frame", row.names = c(NA, 
-507L))

This is to generate the Q-Q plot

example%>% 
  ggplot(aes(sample=v1))  
  stat_qq()   stat_qq_line() 
#   add deciles colour='red' ?

How can I colour as red the deciles?

CodePudding user response：

First create deciles using the ntile function. After that you can color the first decile by filtering that data and color them red. You can use the following code:

library(tidyverse)
example <- example %>% mutate(decile = ntile(v1, 10))

ggplot()   
  geom_qq(example, mapping = aes(sample=example$v1),col=ifelse(example$decile == 1, "red", "black")[order(example$v1)])  
  geom_qq_line(aes(sample=example$v1))

Output:

CodePudding user response：

Thanks to @Quinten answer I could manage to get the decile standing out from the qq points. So I added the ntile()function and calculated the changing points (real deciles) and that's the solution:

library(tidyverse)
example <- example %>% 
  mutate(decile = ntile(v1, 10)) %>% 
  arrange(v1) %>% 
  mutate(difs = decile - lag(decile)) %>% 
  mutate(difs= ifelse(is.na(difs),1,difs))
  
example %>% 
  ggplot(aes(sample=v1))  
  geom_qq(col=ifelse(example$difs == 1, "red", "black"), 
          alpha=ifelse(example$difs == 1, 1, 0.33), 
          size=ifelse(example$difs == 1, 1.5, 1))  
  geom_qq_line()