Sample a proportional number of rows from a grouped data frame

Question

In this data frame:

df <- data.frame(
  Story = c(rep("C", 6), rep("X", 9), rep("A", 15), rep("B",12))
)

I want to randomly sample roughly 33% of all rows in each Story. This proves harder than I thought. This method, for example, using ceiling and slice_sample does not get the desired result:

df %>%
  group_by(Story) %>%
  mutate(ID = row_number()) %>%
  mutate(sample_size = ceiling(n() * 0.33)) %>% 
  slice_sample(n = unique(sample_size))

The desired results has:

• 2 "C"s
• 3 "X"s
• 5 "A"s
• 4 "B"s

Answer 1

What about just prop = 1/3 with slice_sample?

> df %>%
+     slice_sample(prop = 1 / 3, by = Story)
   Story
1      C
2      C
3      X
4      X
5      X
6      A
7      A
8      A
9      A
10     A
11     B
12     B
13     B
14     B

or if you like to use 0.33 and ceiling

> df %>%
+     filter(row_number() %in% sample(n(), ceiling(n() * 0.33)), .by = Story)
   Story
1      C
2      C
3      X
4      X
5      X
6      A
7      A
8      A
9      A
10     A
11     B
12     B
13     B
14     B