Trouble using base-R `rep()` with dplyr

Question

library(tidyverse)
x <- c(0, 20, 30, 58)
n <- 100
df <- data_frame(x, n) %>% 
  distinct() %>%
  filter(x >= 0 & x < n) %>%
  arrange(x) %>%
  bind_rows(data_frame(x = n)) %>%
  mutate(lag_x = lag(x)) %>%
  mutate(y = x - lag_x) %>%
  filter(!is.na(y))

rep(seq_along(df$x), df$y)

The code above works fine. When I try and make the last step pipe into everything it breaks. What's off on my syntax? I want everything to be one long pipe. I think it's impossible because I'd need my whole pipe (the code chunk above) nested inside my rep() call? Is that correct?

library(tidyverse)
x <- c(0, 20, 30, 58)
n <- 100
df <- data_frame(x, n) %>% 
  distinct() %>%
  filter(x >= 0 & x < n) %>%
  arrange(x) %>%
  bind_rows(data_frame(x = n)) %>%
  mutate(lag_x = lag(x)) %>%
  mutate(y = x - lag_x) %>%
  filter(!is.na(y)) %>% 
  rep(seq_along(x), y) %>% 
  print()

Error in function_list[i] : object 'y' not found

Answer 1

As it is outside the mutate/summarise functions, we need to pull or extract it

data_frame(x, n) %>% 
  distinct() %>%
  filter(x >= 0 & x < n) %>%
  arrange(x) %>%
  bind_rows(data_frame(x = n)) %>%
  mutate(lag_x = lag(x)) %>%
  mutate(y = x - lag_x) %>%
  filter(!is.na(y)) %>% 
  {rep(seq_along(.$x), .$y)}

Or it can also be written as

data_frame(x, n) %>% 
 distinct() %>%
 filter(x >= 0 & x < n) %>%
 arrange(x) %>%
 bind_rows(data_frame(x = n)) %>%
 mutate(lag_x = lag(x)) %>%
 mutate(y = x - lag_x) %>%
 filter(!is.na(y)) %>% 
 summarise(n = list(rep(row_number(), y))) %>%
 pull(n)