R/diurnal.R

In Nature40/tRackIT: Process data recorded with automatic radio-tracking stations

#' calculates time of day information
#'
#' @description calculates sunset and sunrise for each data point and the distance to it. For nocturnal species sunrise is +1 day of sunset
#'
#'
#' @author Jannis Gottwald
#'
#' @param data data.frame, data with timestamps
#' @param tcol string, colname of timestamp column
#' @param Lat numeric, reference latitude for sunrise/sunset calculation
#' @param Lon numeric, reference longitude for sunrise/sunset calculation
#' @param tz string, time zone of region ton convert timezone of data to the right format
#'
#'
#' @export
#'
#'





diurnal <- function(data, Lat, Lon, tcol, tz) {
  data$timestamp <- data[, tcol]

  data$timestamp <- lubridate::with_tz(data$timestamp, tzone = tz)

  data$sunrise <- NULL
  data$sunset <- NULL


  data$date <- as.Date(data$timestamp, tz = "CET")


  # sunset time and sunsrise time for all existing dates
  dates <- unique(data$date)
  dates <- dates[!is.na(dates)]
  dates <- format(as.Date(dates), "%Y/%m/%d")
  num <- as.numeric(as.Date(as.character(max(dates)), format = "%Y/%m/%d") - as.Date(as.Date(as.character(min(dates))), format = "%Y/%m/%d"))
  sun <- StreamMetabolism::sunrise.set(Lat, Lon, min(dates), timezone = tz, num.days = num + 1)
  sun$date <- as.Date(sun$sunrise)

  # join to data
  data <- dplyr::inner_join(data, sun)

  data <- as.data.frame(data)

  data$time_to_rise <- as.numeric(difftime(data$timestamp, data$sunrise, units = "hours"))
  data$time_to_set <- as.numeric(difftime(data$timestamp, data$sunset, units = "hours"))


  data$start_datetime <- min(data$timestamp)
  data$stop_datetime <- max(data$timestamp)
  data$year <- lubridate::year(data$date)
  data$ydate <- lubridate::yday(data$date) # Day of the year
  data$month <- lubridate::month(data$date) # Day of the month
  data$t <- data$timestamp - data$start_datetime


  # define the night
  return(data)
}