Process datetimes in UTC (issue #2)

[mstrimas]

Process datetimes in UTC rather than system time zone to ensure returned date times are not offset by 1 day in getSunlightTimes(). Addresses the issue reported at https://github.com/datastorm-open/suncalc/issues/2

[bart1]

Maybe this also relates to #7?

[mfidino]

Adding in here another example of how the current setup does not work (and converting to UTC within getSunlightTimes) fixed the issue. This PR should get merged.

library(lubridate)
library(suncalc)

date = lubridate::ymd("2019-06-04")
lat = 14.96962
lon = 108.1707
tz = "Asia/Ho_Chi_Minh"
data = NULL

example <- suncalc::getSunlightTimes(
  date = lubridate::ymd("2019-06-04"),
  lat = 14.96962,
  lon = 108.1707,
  tz = "Asia/Ho_Chi_Minh"
)

t(
  example
)

              [,1]                 
date          "2019-06-04"         
lat           "14.96962"           
lon           "108.1707"           
solarNoon     "2019-06-05 11:47:08"
nadir         "2019-06-04 23:47:08"
sunrise       "2019-06-05 05:17:58"
sunset        "2019-06-05 18:16:18"
sunriseEnd    "2019-06-05 05:20:23"
sunsetStart   "2019-06-05 18:13:53"
dawn          "2019-06-05 04:54:29"
dusk          "2019-06-05 18:39:47"
nauticalDawn  "2019-06-05 04:26:46"
nauticalDusk  "2019-06-05 19:07:30"
nightEnd      "2019-06-05 03:58:25"
night         "2019-06-05 19:35:51"
goldenHourEnd "2019-06-05 05:48:37"
goldenHour    "2019-06-05 17:45:39"

[bthieurmel]

This path / fix not always working...

See this more complete example :

library(lutz)
library(suncalc)
library(tidyverse)

# default system time zone
Sys.timezone()
#> "America/Los_Angeles"

# example locations along the equator at various longitudes

data_grid <- expand.grid(lat = seq(-90, 90, by = 10), 
                         lon = seq(-180, 180, by = 20))

data_grid$date <- as.Date("2023-06-01")

data <- data_grid %>% 
  mutate(tz = tz_lookup_coords(lat = lat, lon = lon,
                               method = "accurate", warn = FALSE)) 

data$tz <- sapply(data$tz, function(x) strsplit(x, ";")[[1]][1])
# solar noon
res <- data %>% 
  mutate(solar_noon = pmap(list(date = date, lat = lat, lon = lon, tz = tz),
                           getSunlightTimes,
                           keep = "solarNoon") %>% 
           map("solarNoon") %>% 
           map_chr(as.character))

I'm looking to a fully working solution.