CODEWITHSUNDEEP
rggplot2dplyr
Mactilda
Mactilda

Reputation: 393

Label max/min values in a scatterplot in ggplot2 r

I have a large data frame where I want to do a scatterplot where only the max/min values are labeled. 

some_df <- data.frame(
   "Sport" = c(1:5), 
   "avg_height" = c(178, 142, 200, 135, 182), 
   "avg_weight" = c(66, 61, 44, 77, 100))

I have tried: 

library(dplyr)
library(ggplot2)
some_df %>% 
  ggplot(aes(avg_weight, avg_height, label = Sport)) + 
  geom_point(shape = 21) + 
  geom_text(data = subset(avg_height == max(avg_height)))

But get errors telling me that avg_height is not found.

I have also tried with the geom_text

geom_text(aes(label = ifelse(avg_height=max(avg_height), as.character(Sport), '')), 
          hjust=0, vjust=0)

with error for Sport not found.

So I can either label all or none, but with the large data.frame it will be impossible to read. If I can colour only the max/min values it would be fine too. I have experimented with making a new column and trying to join with new variables like below, but it haven't helped me. 

maxw <- some_df %>% summarise_each(Max = max(avg_weight))
maxh <- some_df %>% mutate(summarise(Max = max(avg_height)))

The scatterplot I want is with labels only for the max and min of both avg_heigt and avg_weight.

Upvotes: 0

Views: 5181

Answers (1)

Uwe
Uwe

Reputation: 42544

If I understand correctly, the data points of the extreme values of both avg_weight and avg_weight are supposed to be labeled with the value of Sport:

library(dplyr)
library(ggplot2)
some_df %>% 
  ggplot(aes(avg_weight, avg_height, label = Sport)) + 
  geom_point(shape = 21) + 
  geom_label(data = some_df %>% 
               filter(avg_height %in% range(avg_height) | avg_weight %in% range(avg_weight)),
             nudge_x = 1)

creates

enter image description here

Edit

The OP has asked to label the points with the highest and lowest BMI avg_weight / (avg_height/100)^2 as well:

library(dplyr)
library(ggplot2)
# append BMI column to dataset
some_df <- some_df %>% 
  mutate(bmi = avg_weight / (avg_height/100)^2) 
some_df %>% 
  ggplot(aes(avg_weight, avg_height, label = Sport)) + 
  geom_point(shape = 21) + 
  geom_label(data = some_df %>% 
               filter(
                 avg_height %in% range(avg_height) | 
                   avg_weight %in% range(avg_weight) |
                   bmi %in% range(bmi)
               ),
             nudge_x = 1)

The resulting chart is the same as above.

Upvotes: 4

Related Questions