add the historical maximum temperature data

R/data.R

@@ -1,8 +1,11 @@
 #' Australia climate data
 #'
-#' Daily measure on precipitation (\code{prcp}), maximum temperature
-#' (\code{tmax}), and minimum temperature (\code{tmin}) in 2020 for
-#' 639 stations.
+#' \code{climate_aus}: daily measure on precipitation (\code{prcp}),
+#' maximum temperature (\code{tmax}), and minimum temperature (\code{tmin})
+#' in 2020 for 639 stations. \code{historical_tmax}: daily maximum temperature
+#' (\code{tmax}) for 75 stations in Victoria and New South Wales for two
+#' periods: 1971-1975 and 2016-2020.
+#'
 #' @details
 #' \describe{
 #'   \item{id}{station ID, "ASN000" are international paddings, the next
@@ -16,13 +19,17 @@
 #'   \item{elev}{elevation of the stations}
 #'   \item{name}{station name}
 #'   \item{wmo_id}{the world meteorological organisation (WMO) station number}
-#'   \item{ts}{a list-column that nests all the temporal variables:
-#'   date, prcp, tmax, and tmin}
+#'   \item{ts}{For \code{climate_aus}: date, prcp, tmax, and tmin, for
+#'   \code{historical_tmax}: date and tmax}
 #' }
+#' @rdname climate-data
 #' @examples
 #' climate_aus %>% face_temporal() %>% face_spatial()
 "climate_aus"
 
+#' @rdname climate-data
+"historical_tmax"
+
 #' Toy climate data
 #'
 #' Daily measure (2020-01-01 to 2020-01-10) on precipitation (prcp),

data-raw/tmax-hist.R

@@ -1,38 +1,39 @@
-## code to prepare `DATASET` dataset goes here
-library(rnoaa)
 library(tidyverse)
 library(cubble)
-all_stations <- ghcnd_stations() %>%
-  filter(str_starts(id, "ASN")) %>%
-  filter(last_year >= 2020) %>%
-  mutate(wmo_id = as.numeric(wmo_id),
-         name = str_to_lower(name)) %>%
-  select(-state, -gsn_flag)
+all_stations <- rnoaa::ghcnd_stations() |>
+  filter(str_starts(id, "ASN")) |> # Australian stations start wiht "ASN"
+  filter(last_year >= 2020) |>
+  mutate(wmo_id = as.numeric(wmo_id), name = str_to_lower(name)) |>
+  select(-state, -gsn_flag) |>
+  select(id, longitude, latitude, elevation, name,
+         wmo_id, element, first_year, last_year) |>
+  rename(long = longitude, lat = latitude, elev = elevation)
 
-cand <- all_stations %>%
-  filter(element == "TMAX",
-         first_year < 1970,
-         !is.na(wmo_id))
+tmax_stations <- all_stations |>
+  filter(element == "TMAX", first_year < 1970, !is.na(wmo_id)) %>%
+  filter(between(as.numeric(stringr::str_sub(id, 7, 8)), 46, 90)) %>%
+  filter(!id %in% c("ASN00047048", "ASN00085279", "ASN00085298")) %>%
+  select(-c(element:last_year))
 
-historical_tmax <- cand %>%
-  rowwise() %>%
-  mutate(ts = list(meteo_pull_monitors(
-    monitors = id, var = "TMAX",
-    date_min = glue::glue("{first_year}-01-01"),
-    date_max = glue::glue("{last_year}-12-31")) |>
-      select(-id))) %>%
-  as_cubble(index = date, key = id, coords = c(longitude, latitude))
+tmax71 <- rnoaa::meteo_pull_monitors(
+  monitors = tmax_stations$id, var = "TMAX",
+  date_min = glue::glue("1971-01-01"),
+  date_max = glue::glue("1975-12-31")
+)
 
-tmax_hist <- historical_tmax |>
-  face_temporal() |>
-  mutate(tmax = tmax / 10) |>
-  filter(between(lubridate::year(date), 1970, 1975) |
-           lubridate::year(date) > 2015) |>
-  mutate(yearmonth = tsibble::yearmonth(date)) |>
-  group_by(yearmonth) |>
-  summarise(tmax = mean(tmax, na.rm = TRUE)) |>
-  face_spatial() |>
-  select(-c(element:last_year)) |>
-  filter(stringr::str_sub(id, 7, 8) >= 76)
+tmax16 <- rnoaa::meteo_pull_monitors(
+  monitors = tmax_stations$id, var = "TMAX",
+  date_min = glue::glue("2016-01-01"),
+  date_max = glue::glue("2020-12-31")
+)
 
-usethis::use_data(tmax_hist, overwrite = TRUE,  compress = "gzip")
+tmax <- tmax71 %>%
+  bind_rows(tmax16) %>%
+  arrange(id, date) %>%
+  mutate(tmax = tmax /10)
+
+historical_tmax <- make_cubble(
+  spatial = tmax_stations, temporal = tmax,
+  key = id, index = date, coords = c(long, lat))
+
+usethis::use_data(historical_tmax, overwrite = TRUE)