Find average temperature from a range of datetime for each day in dataframe

July 26, 2024

This is a subset of the dataframe I have:

structure(list(name = c("waldorf", "waldorf", "waldorf", "waldorf", 
"waldorf", "waldorf", "waldorf", "waldorf", "waldorf", "waldorf", 
"waldorf", "waldorf", "waldorf", "waldorf", "waldorf", "waldorf", 
"waldorf", "waldorf", "waldorf", "waldorf", "waldorf", "waldorf", 
"waldorf", "waldorf"), date = structure(c(1559347200, 1559347200, 
1559347200, 1559347200, 1559347200, 1559347200, 1559347200, 1559347200, 
1559347200, 1559347200, 1559347200, 1559347200, 1559347200, 1559347200, 
1559347200, 1559347200, 1559347200, 1559347200, 1559347200, 1559347200, 
1559347200, 1559347200, 1559347200, 1559347200), tzone = "UTC", class = c("POSIXct", 
"POSIXt")), time = structure(c(0, 3600, 7200, 10800, 14400, 18000, 
21600, 25200, 28800, 32400, 36000, 39600, 43200, 46800, 50400, 
54000, 57600, 61200, 64800, 68400, 72000, 75600, 79200, 82800
), class = c("hms", "difftime"), units = "secs"), datetime = structure(c(1559347200, 
1559350800, 1559354400, 1559358000, 1559361600, 1559365200, 1559368800, 
1559372400, 1559376000, 1559379600, 1559383200, 1559386800, 1559390400, 
1559394000, 1559397600, 1559401200, 1559404800, 1559408400, 1559412000, 
1559415600, 1559419200, 1559422800, 1559426400, 1559430000), tzone = "UTC", class = c("POSIXct", 
"POSIXt")), temp = c(72.9, 70.6, 69.7, 69.6, 68, 68.1, 67.9, 
67.8, 69.6, 71.9, 75.6, 78.3, 80.1, 80.8, 82.8, 83.5, 83.3, 82.9, 
81.5, 79.4, 77.6, 72.4, 70.6, 69.8)), row.names = c(NA, -24L), class = c("tbl_df", 
"tbl", "data.frame"))

The expanded data frame has hourly temperature data for 5 years worth of days. How can I use the datetime or the other columns to calculate the average temperature from 1pm to 5pm for each day in the data frame? Can I do this in a dplyr string?

>Solution :

Here are two variations. The first filters the data to times from 1pm (13 in 24h time) to 4:59:59 and then summarizes to give one row per date.

library(dplyr); library(lubridate)
df |>
  filter(hour(datetime) %in% 13:16) |>
  summarize(avg_temp = mean(temp), .by = date)

# A tibble: 1 × 2
  date                avg_temp
  <dttm>                 <dbl>
1 2019-06-01 00:00:00     82.6

Or we could could keep every row and add a new column that just references the times from 1-5pm for that date.

df |>
  mutate(avg_temp = mean(temp[hour(datetime) %in% 13:16]), .by = date)

   name    date                time   datetime             temp avg_temp
   <chr>   <dttm>              <time> <dttm>              <dbl>    <dbl>
 1 waldorf 2019-06-01 00:00:00 00:00  2019-06-01 00:00:00  72.9     82.6
 2 waldorf 2019-06-01 00:00:00 01:00  2019-06-01 01:00:00  70.6     82.6
 3 waldorf 2019-06-01 00:00:00 02:00  2019-06-01 02:00:00  69.7     82.6
 4 waldorf 2019-06-01 00:00:00 03:00  2019-06-01 03:00:00  69.6     82.6
 5 waldorf 2019-06-01 00:00:00 04:00  2019-06-01 04:00:00  68       82.6
 6 waldorf 2019-06-01 00:00:00 05:00  2019-06-01 05:00:00  68.1     82.6
 7 waldorf 2019-06-01 00:00:00 06:00  2019-06-01 06:00:00  67.9     82.6
 8 waldorf 2019-06-01 00:00:00 07:00  2019-06-01 07:00:00  67.8     82.6
 9 waldorf 2019-06-01 00:00:00 08:00  2019-06-01 08:00:00  69.6     82.6
10 waldorf 2019-06-01 00:00:00 09:00  2019-06-01 09:00:00  71.9     82.6
# ℹ 14 more rows