CODEWITHSUNDEEP
r
Alexey
Alexey

Reputation: 1

Calculating looped averages

I have a table containing dates within a single year and their corresponding average temperatures (mean_temp) for each day. My objective is to compute the average temperature for the next 30 days following each given date. However, I encountered difficulty in handling the last 30 observations.

So the data looks similar to this:

temp_data = tibble(
  day_year = c(1:366),
  temp = rnorm(366,5,2) 
  )

My initial attempt (below) resulted in either "error in temp_data$temp[1:(day_year - 335)]: ! only 0's may be mixed with negative subscripts" or computing only one value for all last observations if i take absolute values of (day_year - 335).

temp_data |> 
  mutate(
    temp_next_30d = if_else(
      day_year < 336,
      rollapply(temp, width = 30, FUN = mean, align = "right", fill = NA, partial = TRUE),
      mean(temp_data$temp[day_year:366]) + mean(temp_data$temp[1:(day_year-335)])
    )
  )

Are there any elegant solutions to address this?

Upvotes: 0

Views: 30

Answers (1)

Edward
Edward

Reputation: 19169

One option is to add the first 29 days to the data, calculate the rolling mean, and then remove those rows. And I think you want align='left`.

temp_data2 <- temp_data |> 
  bind_rows(slice(temp_data, 1:29)) |>
  mutate(
    temp_next_30d = zoo::rollapply(temp, width = 30, FUN = mean, 
                                   align = "left", fill = NA, partial = TRUE)) |>
  slice(1:366)

temp_data2
# A tibble: 366 × 3
   day_year  temp temp_next_30d
      <int> <dbl>         <dbl>
 1        1  4.30          4.51
 2        2  5.31          4.47
 3        3  6.04          4.37
 4        4  2.36          4.31
 5        5  3.83          4.44
 6        6  6.04          4.53
 7        7  4.48          4.50
 8        8  3.38          4.55
 9        9  5.04          4.71
10       10  4.30          4.72

tail(temp_data2)
# A tibble: 6 × 3
  day_year  temp temp_next_30d
     <int> <dbl>         <dbl>
1      361  4.91          4.61
2      362  4.92          4.58
3      363  4.44          4.58
4      364  4.26          4.58
5      365  5.78          4.63
6      366  8.17          4.64

Upvotes: 1

Related Questions