Should spread have digits argument?

[hadley]

To avoid mutate here:

library(tidyr)
library(dplyr)

prob <- seq(0, 1, length = 10)
mtcars %>%
  group_by(cyl) %>%
  summarise(q = list(quantile(mpg, prob)), p = list(prob)) %>%
  unnest() %>%
  mutate(p = format(p, digits = 2)) %>%
  spread(p, q, sep = "_")

[hadley]

Now handled via pivot spec: if you want greater control over column names, you can generate the col_names yourself