Create new column with frequency of values

Question

This is similar to my previous questions posted: Create new column with distinct character values

But I also wanted some additional information.

df:

ID <- c(1,1,1,1,1,1,1,2,2,2,2,2)
color <- c("red","red","red","blue","green","green","blue",
           "yellow","yellow","red","blue","green")
df <- data.frame(ID,color)

   ID  color
1   1    red
2   1    red
3   1    red
4   1   blue
5   1  green
6   1  green
7   1   blue
8   2 yellow
9   2 yellow
10  2    red
11  2   blue
12  2  green

Creating n_distinct_color (number of distinct colors each ID has):

df %>% 
  group_by(ID) %>% 
  distinct(color, .keep_all = T) %>% 
  mutate(n_distinct_color = n(), .after = ID) %>% 
  ungroup()

# A tibble: 7 × 3
     ID n_distinct_color color 
  <dbl>            <int> <chr> 
1     1                3 red   
2     1                3 blue  
3     1                3 green 
4     2                4 yellow
5     2                4 red   
6     2                4 blue  
7     2                4 green

Now I want to create:

1. new "Frequency" column that shows how many times each color appears for each ID (From original df, ID 1 has 3 red, 2 blue, 2 green, etc)
2. new "most frequent color" column that shows which color is the most frequent for each ID. (From original df, most frequent color for ID1 is red, for ID2 is yellow.)

     ID n_distinct_color color    frequency_of_color   most_frequent_color 
  <dbl>            <int> <chr>    <int>                <chr>
1     1                3 red      3                    red
2     1                3 blue     2                    red
3     1                3 green    2                    red
4     2                4 yellow   2                    yellow
5     2                4 red      1                    yellow
6     2                4 blue     1                    yellow
7     2                4 green    1                    yellow

Also, what if there's a case where there are 2 colors with the same frequency (ie, ID 2's most frequent color are yellow and red, how will the data table be like?)

df_new:

ID <- c(1,1,1,1,1,1,1,2,2,2,2,2,2)
color <- c("red","red","red","blue","green","green","blue",
           "yellow","yellow","red","blue","green","red")
df_new <- data.frame(ID,color)

   ID  color
1   1    red
2   1    red
3   1    red
4   1   blue
5   1  green
6   1  green
7   1   blue
8   2 yellow
9   2 yellow
10  2    red
11  2   blue
12  2  green
13  2    red

I would appreciate all the help there is! Thanks!!!

Answer 1

With a series of mutate and summarise you can achieve your goal. In case of ties, here [1] means the first tied color is chosen:

library(dplyr) #1.1.0 or above required
df %>% 
  mutate(n_distinct = n_distinct(color), .by = ID) %>% 
  summarise(frequency = n(), .by = c(ID, n_distinct, color)) %>% 
  mutate(most_frequent = color[which.max(frequency)[1]], .by = ID)

output

  ID n_distinct  color frequency most_frequent
1  1          3    red         3           red
2  1          3   blue         2           red
3  1          3  green         2           red
4  2          4 yellow         2        yellow
5  2          4    red         2        yellow
6  2          4   blue         1        yellow
7  2          4  green         1        yellow