Subset rows using slice_max() - how to use it in R?

I have to create a partial data set containing only the 20 days with the highest daily mean air temperature values for each year. My dataset looks like this:

date	mean
1997-07-15	27.05292
1997-07-17	26.86542
1997-06-21	26.10958
1997-07-16	26.05833
1997-07-14	26.02500
1997-06-25	25.80125
1997-07-18	25.36208
1997-06-22	25.18875
1997-06-29	24.72333
1997-06-30	24.71000

...

I tried to use the code bellow, but this one only filters the maximum from every year and creates a dataframe with 20 rows - but I need the Top 20 mean values from every year (1997 – 2010). I use the class data.frame btw. I would be so grateful if anyone can help me, I just can't figure it out!

top_20_per_year <- daily_mean_temp_sorted %>%
  slice_max(mean, n = 20) %>%

Solution

Example taking the top 2 mean values by year:

library(tidyverse)

df <- tribble(
  ~date, ~mean,
  "1997-07-15", 27.05292,
  "1997-07-17", 26.86542,
  "1997-06-21", 26.10958,
  "1997-07-16", 26.05833,
  "1997-07-14", 26.02500,
  "1998-06-25", 25.80125,
  "1998-07-18", 25.36208,
  "1998-06-22", 25.18875,
  "1998-06-29", 24.72333,
  "1998-06-30", 24.71000
)

df |> 
  mutate(date = ymd(date), year = year(date)) |> 
  slice_max(n = 2, order_by = mean, by = year)
#> # A tibble: 4 × 3
#>   date        mean  year
#>   <date>     <dbl> <dbl>
#> 1 1997-07-15  27.1  1997
#> 2 1997-07-17  26.9  1997
#> 3 1998-06-25  25.8  1998
#> 4 1998-07-18  25.4  1998

^{Created on 2024-04-29 with reprex v2.1.0}