Added transit and tests

DESCRIPTION

@@ -14,10 +14,15 @@ Encoding: UTF-8
 Roxygen: list(markdown = TRUE)
 RoxygenNote: 7.3.2
 Imports: 
+    archive,
+    cli,
     data.table,
     dplyr,
     httr2,
     jsonlite,
+    lubridate,
     sf,
-    stringr
+    stringr,
+    tibble,
+    tidyr
 URL: https://samherniman.github.io/autocruller/

NAMESPACE

@@ -1,3 +1,7 @@
 # Generated by roxygen2: do not edit by hand
 
 export(ac_get_co2)
+export(ac_get_co2_download)
+export(ac_get_co2_transit)
+export(ac_get_co2_web)
+export(ac_unnest_longer)

R/ac_get_co2.R

@@ -1,44 +1,43 @@
 #' Get data from indoorco2map.org
+#' 
+#' @param x Character specifying which dataset to download one of c("web", "download", "transit")
 #'
 #' @returns A dataframe with up to date co2 measurements
 #' @export
 #'
 #' @examples
 #' \dontrun{
+#' # Get building data with one of three options
 #' ac_df <- ac_get_co2()
+#' ac_df <- ac_get_co2("web")
+#' ac_df <- ac_get_co2("download")
+#' 
+#' # Get transit data
+#' ac_df <- ac_get_co2("transit")
 #' }
 #'
-ac_get_co2 <- function() {
-  measurements <- .id <- startTime <- NULL
+ac_get_co2 <- function(x = "web") {
 
-  icm_response <-
-    httr2::request("https://indoorco2map.com/chartdata/IndoorCO2MapData.json.gz") |>
-    httr2::req_perform() |>
-    httr2::resp_body_json()
+  if (!(x %in% c("web", "download", "transit"))) {
+    cli::cli_abort(c(
+      "Error while getting co2 data:",
+      "x must be either \"web\", \"download\", or \"transit\"",
+      "if you want building data, set x to \"web\"",
+      "if you want transit data, set x to \"transit\""
+    ))
+  }
+
+  if (x == "web") {
+    return(ac_get_co2_web())
+  }
 
-  icm_response <-
-    data.table::rbindlist(icm_response, idcol = TRUE) |>
-    suppressWarnings()
+  if (x == "download") {
+    return(ac_get_co2_download())
+  }
 
-  j_lst <- lapply(icm_response$measurements, jsonlite::fromJSON)
-  j_lst <- lapply(j_lst, co2_to_numeric)
-  j_lst <- data.table::rbindlist(j_lst, idcol = TRUE)
-
-  ac_df <- icm_response |>
-    dplyr::select(-measurements) |>
-    dplyr::left_join(j_lst, by = dplyr::join_by(.id)) |>
-    dplyr::mutate(
-      # remove the last few digits from the unix time because they are in milliseconds and not needed
-      date = stringr::str_sub(startTime, end = -4) |>
-        as.numeric() |>
-        as.POSIXct()
-    ) |>
-    sf::st_as_sf(
-      coords = c("lon", "lat"),
-      crs = sf::st_crs(4326)
-    )
-
-  return(ac_df)
+  if (x == "transit") {
+    return(ac_get_co2_transit())
+  }
 }
 
 #' Convert the co2 character array to a numeric list

R/ac_get_co2_download.R

@@ -0,0 +1,39 @@
+
+#' Get data from indoorco2map.org - "download version"
+#'
+#' @returns A dataframe with co2 building measurements
+#'
+#' @details
+#' This function will download the version of building 
+#' data that is available using the "download" button on 
+#' the website
+#' 
+#' 
+#' @export
+#' @examples
+#' 
+#' \dontrun{
+#' ac_df <- ac_get_co2_download()
+#' }
+ac_get_co2_download <- function() {
+  co2readingsAvg <- startOfMeasurement <- NULL
+  icm_response <-
+    archive::archive_extract(
+      "https://indoorco2map.com/chartdata/IndoorCO2MapData.zip",
+      files = 3
+    ) |> 
+  jsonlite::fromJSON() |> 
+  tibble::rowid_to_column("obs_number") |> 
+  dplyr::mutate(
+    date = lubridate::as_datetime(startOfMeasurement),
+    day = lubridate::date(date)
+  ) |> 
+    sf::st_as_sf(
+      coords = c("longitude", "latitude"),
+      crs = sf::st_crs(4326)
+    )
+  unlink('indoorco2mapData.json')
+
+  return(icm_response)
+}
+

R/ac_get_co2_transit.R

@@ -0,0 +1,93 @@
+#' Get transit data from indoorco2map.org
+#'
+#' @returns A dataframe with up to date co2 transit measurements
+#'
+#' @export
+#' @examples
+#' \dontrun{
+#' ac_df <- ac_get_co2_transit()
+#' }
+ac_get_co2_transit <- function() {
+    icm_response <-
+      httr2::request("https://rkhby3mvq3.execute-api.eu-central-1.amazonaws.com/GetTransportCO2Data") |>
+      httr2::req_perform() |>
+      httr2::resp_body_json() |> 
+    unlist() |> 
+    lapply(jsonlite::fromJSON)
+
+  j_df <- do.call(rbind, icm_response) |> 
+    tibble::rowid_to_column("uid") 
+
+  j_df <- lapply(1:nrow(j_df), \(x) json_to_df(j_df[x,]))
+  j_df <- 
+    do.call(rbind, j_df) |>
+    sf::st_as_sf(
+      coords = c("long_first", "lat_first"), 
+      crs = sf::st_crs(4326)
+    ) |>
+    sf::st_make_valid() 
+
+  return(j_df)
+}
+
+#' Turn a co2 json into a dataframe
+#'
+#' @param x A json 
+#'
+#' @returns a dataframe
+json_to_df <- function(x) {
+  timestampArray <- 
+    co2Array <- 
+    longitudeArray <- 
+    latitudeArray <- 
+    startTime <- 
+    uid <- 
+    ppmAvg <- 
+    NULL
+  x |> 
+    dplyr::mutate(
+      timestampArray = split_to_numeric(timestampArray) |> range01() |> paste(collapse = ";"),
+      # mod_loess = list(train_loess(co2_vec = co2Array, time_vec = timestampArray)),
+      co2Array = stringr::str_replace_all(co2Array, ";", ","),
+      longitudeArray = stringr::str_remove_all(longitudeArray, "(0;|0$)") |> 
+        stringr::str_replace_all(",", "\\.") |> 
+        stringr::str_replace_all(";", ","),
+      latitudeArray = stringr::str_remove_all(latitudeArray, "(0;|0$)") |>
+        stringr::str_replace_all(",", "\\.") |> 
+        stringr::str_replace_all(";", ","),
+      date = stringr::str_sub(startTime, end = -4) |>
+        as.numeric() |>
+        as.POSIXct()
+    ) |> 
+    tidyr::separate_longer_delim(cols = c(co2Array), delim = ",") |> 
+    dplyr::distinct() |> 
+    dplyr::mutate(
+      long_first = stringr::str_split_i(longitudeArray, ",", 1) |> as.numeric(),
+      lat_first = stringr::str_split_i(latitudeArray, ",", 1) |> as.numeric()
+    ) |> 
+    # sf::st_as_sf(coords = c("long_first", "lat_first"), crs = sf::st_crs(4326)) |> 
+    dplyr::group_by(uid) |> 
+    tibble::rowid_to_column("tsa") |> 
+    dplyr::select(-ppmAvg) |> 
+    tidyr::drop_na() 
+}
+
+#' Create a vector of values evenly spaced between 0 and 1
+#'
+#' @param x vector of values
+#' @param ... other parameters passed to min() or max()
+#'
+#' @returns vector of values evenly spaced between 0 and 1 the same length as x
+range01 <- function(x, ...){(x - min(x, ...)) / (max(x, ...) - min(x, ...))}
+
+#' Split a string seperated by ";" into numbers
+#'
+#' @param x a string of numbers seperated by ";"
+#'
+#' @returns a numeric vector
+split_to_numeric <- function(x) {
+  x |> 
+    strsplit(split = ";") |>
+    unlist() |> 
+    as.numeric()
+}

R/ac_get_co2_web.R

@@ -0,0 +1,64 @@
+#' Get data from indoorco2map.org
+#'
+#' @returns A dataframe with up to date co2 measurements - "web version"
+#' @export
+#' 
+#' @details
+#' This function will download the version of building 
+#' data that is displayed on the website map 
+#' 
+#'
+#' @examples
+#' \dontrun{
+#' ac_df <- ac_get_co2_web()
+#' }
+#'
+ac_get_co2_web <- function() {
+  measurements <- .id <- startTime <- obs_number <- co2array <- NULL
+
+  icm_response <-
+    httr2::request("https://indoorco2map.com/chartdata/IndoorCO2MapData.json.gz") |>
+    httr2::req_perform() |>
+    httr2::resp_body_json()
+
+  icm_response <-
+    data.table::rbindlist(icm_response, idcol = "obs_number") |>
+    suppressWarnings()
+
+  j_lst <- lapply(icm_response$measurements, jsonlite::fromJSON)
+  j_lst <- lapply(j_lst, co2_to_numeric)
+  j_lst <- data.table::rbindlist(j_lst, idcol = "obs_number")
+
+  ac_df <- icm_response |>
+    dplyr::select(-measurements) |>
+    dplyr::left_join(j_lst, by = dplyr::join_by(obs_number)) |>
+    dplyr::mutate(
+      # remove the last few digits from the unix time because they are in milliseconds and not needed
+      date = stringr::str_sub(startTime, end = -4) |>
+        as.numeric() |>
+        as.POSIXct()
+    ) |>
+    dplyr::select(-co2array) |> 
+    sf::st_as_sf(
+      coords = c("lon", "lat"),
+      crs = sf::st_crs(4326)
+    )
+
+  return(ac_df)
+}
+
+#' Convert the co2 character array to a numeric list
+#'
+#' @param x a list containing measurements from json
+#'
+#' @returns list of co2 records
+#'
+co2_to_numeric <- function(x) {
+  x$co2readings <-
+    x$co2array |>
+    stringr::str_split(";", simplify = TRUE) |>
+    as.numeric() |>
+    list()
+
+  return(x)
+}

R/ac_unnest_longer.R

@@ -0,0 +1,27 @@
+#' Unnest a co2 df list-column into rows
+#'
+#' @param x a dataframe from ac_get_co2() 
+#'
+#' @returns a dataframe in long format
+#'
+#' @export
+#' @examples
+#' ac_df <- ac_get_co2("download") |> 
+#' ac_unnest_longer()
+ac_unnest_longer <- function(x) {
+  co2readings <- obs_number <- offset <- interval <- reading_index <- NULL
+  x_df <- x |> tidyr::unnest_longer(co2readings)
+
+  if (all(c("offset", "interval") %in% names(x))) {
+    x_df <- x_df |>
+    dplyr::group_by(obs_number) |> 
+    dplyr::mutate(
+      reading_index = dplyr::row_number(),
+      date_time = lubridate::as_datetime(date) +
+        lubridate::dminutes(offset) + 
+        lubridate::dminutes(interval * reading_index)
+    )
+  }
+
+  return(x_df)
+}

README.Rmd

@@ -58,7 +58,7 @@ Download the latest data with `ac_get_co2()`
 ```{r getco2}
 library(autocruller)
 
-ac_df <- ac_get_co2()
+ac_df <- ac_get_co2("download")
 ```
 
 That will give you a dataframe with all the current CO^2^ measurements. 
@@ -67,10 +67,10 @@ That will give you a dataframe with all the current CO^2^ measurements.
 dplyr::glimpse(ac_df)
 ```
 
-If you want the co2arrays to be in long format, you can do this:
+If you want the co2readingss to be in long format, you can do this:
 
 ```{r}
-ac_df_long <- tidyr::unnest_longer(ac_df, co2array)
+ac_df_long <- ac_unnest_longer(ac_df)
 ```
 
 
@@ -91,11 +91,11 @@ types_c <- c(
   "music", "nightclub", "social_centre", "concert_hall"
 )
 ac_df |>
-  dplyr::filter(osmtag %in% types_c) |>
+  dplyr::filter(osmTag %in% types_c) |>
   tidyplot(
-    x = ventilation,
-    y = ppmavg,
-    color = ppmavg
+    x = ventilationSystem,
+    y = co2readingsAvg,
+    color = co2readingsAvg
   ) |>
   add_data_points_beeswarm(size = 3) |>
   adjust_y_axis(transform = "log2") |>
@@ -112,21 +112,18 @@ And track things over time with the long format data
 
 ```{r}
 ac_df_long |>
-  dplyr::filter(.id == 127) |>
-  dplyr::mutate(
-    measurement_sequence = dplyr::row_number()
-  ) |>
+  dplyr::filter(obs_number == 458) |>
   tidyplot(
-    x = measurement_sequence,
-    y = co2array,
-    color = co2array
+    x = reading_index,
+    y = co2readings,
+    color = co2readings
   ) |>
   add_data_points() |>
   adjust_size(width = NA, height = NA, unit = "cm") |>
   adjust_font(fontsize = 16) |>
   adjust_x_axis_title("Time") |>
   adjust_y_axis_title("CO2 ppm") |>
   remove_legend() |>
-  adjust_title("CO2 measurements at recording 127") |>
+  adjust_title("CO2 measurements at recording 458") |>
   adjust_caption("Data from indoorCO2map.com")
 ```