In BenjaminLouis/refele: refele : a package for the EFELE project

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(refele)

Summarise hourly climatic data towards daily climatic data

Climatic data can be obtained at hourly intervals but need to be used at daily scale. Function summarise_daily allows to summarise hourly climatic data towards daily climatic data either by calculating a cumulative sum of the data aver each day (e.g. for precipitation data) or by calculating a daily mean value of the data (e.g. for temperature data).

You can see how ti use the function with :

?sumarise_daily

Here is an example with 3-hours interval climatic dataset :

library(dplyr, warn.conflicts = FALSE)
library(lubridate, warn.conflicts = FALSE)
# Convert date time column to date time class
hourlydat <- meteoClermont %>%
  mutate(date = ymd_hms(date))
# Run the function
dailydat <- summarise_daily(.data = hourlydat, .timevar = date, .cumul = precipitation , 
                .average = wind_dir:pres)

summarise_daily accept a dataframe (.data) with a column corresponding to the date and time of data measurement (.timevar). If the classe of .timevar is not a R Date-Time class (POSIXct), the function will try to parse it but you will get a warnings that you should do it on your own before, to be sure that the good format is used. In this example, function lubridate::ymd_hms is used in order to do it.

Argument .cumul is used to give the name of columns that need to be summed and .average is used to the column that need to be averaged. Here, all columns betweens wind_dir and pres are averaged and only the column precipitation is summed.

We can see the graph of wind direction measured at hourly intervals (black dots) and its daily mean (blue line) :

library(ggplot2)
dat <- left_join(
  select(hourlydat, date, wind_dir_hourly = wind_dir) %>% 
    mutate(the_days = floor_date(date, unit = "day")),
  select(dailydat, the_days, wind_dir_daily = wind_dir),
  by = "the_days"
)
ggplot(dat) +
  geom_point(aes(x = date, y = wind_dir_hourly), shape = 20, colour = "black") +
  geom_line(aes(x = date, y = wind_dir_daily), colour = "blue") +
  labs(x = "Date", y = "Wind direction")