fars_packagebuild: FARS Data Tools

# Tyler Byers
# Coursera "Building R Packages"
# Week 2 Assigment -- Documenting Functions
# Nov 23 2016

#' Read FARS Data.
#'
#' \code{fars_read} reads FARS data into the environment.
#'
#'  This function reads data from the  US National Highway Traffic Safety
#'   Administration's Fatality Analysis Reporting System (FARS), given a
#'   filename for the data. It returns a tibble of the data. For this function
#'   to work properly, a filename pointing to an existing file must be given.
#'
#' @param filename A character string giving the filename of the FARS data.
#' @return This function returns a tibble containing the FARS data. If an
#'   incorrect filename is entered the function will stop.
#' @examples
#' full_filename <- system.file('extdata', 'accident_2013.csv.bz2',
#'                              package = 'farsdata')
#' fars_read(filename = full_filename)
#'
#' \dontrun{
#' fars_read(filename = 'filedoesnotexist')
#' }
#'
#' @export
fars_read <- function(filename) {
  if(!file.exists(filename))
    stop("file '", filename, "' does not exist")
  data <- suppressMessages({
    readr::read_csv(filename, progress = FALSE)
  })
  dplyr::tbl_df(data)
}


#' Make FARS Filename.
#'
#' \code{make_filename} makes a properly formatted FARS filename given a year
#'   as input.
#'
#' This function takes a year as input and produces a valid FARS filename. This
#'   function is "dumb" in that it will not check whether the file
#'   actually exists, or if the year is reasonable -- you could enter any number
#'   you like.
#'
#' @param year An integer, or a string or numeric that can be coerced to a string,
#'   of the year of interest.
#' @return this function returns a string that is the proper FARS data
#'    filename for the given year.  Will return a filename with NA for the
#'    year slot in the name if the year parameter cannot be coerced to an
#'    integer.
#' @examples
#' make_filename(year = '2013')
#' make_filename(year = 2013)
#'
#' \dontrun{
#' make_filename(year = 'two thousand thirteen') #  error
#' }
#'
#'
#' @export
make_filename <- function(year) {
  year <- as.integer(year)
  filename <- sprintf("accident_%d.csv.bz2", year)
  full_filename <- system.file('extdata', filename, package = 'farsdata')
  full_filename
}

#' Read FARS files for one or more years.
#'
#' \code{fars_read_years} produces a list of tibbles of FARS data, given an
#'   input vector of years.
#'
#' This function takes a vector of years and produces a list of tibbles,
#'   where each tibble is that year's FARS file year and MONTH observations.
#'   This is a simple function that strips all useful data out of the FARS
#'   tables and produces a completely useless tibble, but is meant for
#'   practice in the Coursera course.  For this function to work, valid
#'   years should be entered. Invalid years will have a NULL entry in the
#'   returned list.
#'
#' @param years Vector of years' FARS files to open.  Vector members must be
#'    an integer, or a string or numeric that can be coerced to a string,
#'    of the year of interest.
#' @importFrom magrittr "%>%"
#' @return This function returns a list of tibbles, where each tibble contains
#'   containing the year and month from the observations in the corresponding
#'   year's FARS data.  If an invalid year is given, the corresponding
#'   list will be NULL.
#' @examples
#' fars_read_years(years = c(2013, 2014, 2015))
#' fars_read_years(years = 2013)
#'
#' \dontrun{
#' fars_read_years(years = 2000) # error
#' }
#'
#' @export
fars_read_years <- function(years) {
  lapply(years, function(year) {
    file <- make_filename(year)
    tryCatch({
      dat <- fars_read(file)
      dplyr::mutate(dat, year = year) %>%
        dplyr::select_(~ MONTH, ~ year)
    }, error = function(e) {
      warning("invalid year: ", year)
      return(NULL)
    })
  })
}


#' Produce a Summary of FARS Files.
#'
#' \code{fars_summarize_years} produces a summary tibble of FARS years and
#'   months given a vector of years.
#'
#' This function takes a vector of years, pulls the FARS data for
#'   those years, and then produces a summary tibble. The summary tibble shows
#'   the number of observations for each month/year combination for the
#'   extracted FARS data. For this function to work properly, the years must
#'   be years with valid data.
#'
#' @param years Vector of years' FARS files to open.  Vector members must be
#'    an integer, or a string or numeric that can be coerced to a string.
#' @return This function returns tibble where the first column is the month,
#'   the second and following columns are the requested years, and the
#'   rows for the year columns are the number of FARS observations for
#'   that month/year combination.  The returned columns are only for years
#'   with valid FARS data. If no valid years are found, the function
#'   well error out.
#' @examples
#' fars_summarize_years(years = c(2013, 2014, 2015))
#' fars_summarize_years(years = 2013)
#'
#' \dontrun{
#' fars_summarize_years(years = 2000)
#' }
#'
#' @export
fars_summarize_years <- function(years) {
  dat_list <- fars_read_years(years)
  dplyr::bind_rows(dat_list) %>%
    dplyr::group_by_(~ year, ~ MONTH) %>%
    dplyr::summarize_(n = ~ n()) %>%
    tidyr::spread_('year', 'n')
}


#' Map State Motor Vehicle Fatalities.
#'
#' \code{fars_map_state} maps state motor vehicle fatalities given a year and
#'   state id number.
#'
#' This function takes a state number and a year, and draws
#'   a state outline with dots to represent the location of motor vehicle
#'   fatalities for that year.  This function will throw an error if an invalid
#'   state number is chosen or the chosen year's data does not exist.
#'
#'   You must have library(mapdata) loaded in your namespace for this to work.
#'
#' @param state.num Numerical code for US state.
#' @param year  An integer, or a string or numeric that can be coerced to a string,
#'   of the year of interest.
#' @return NULL
#' @examples
#' library(mapdata)
#' fars_map_state(12, 2014)
#' fars_map_state(36, 2014)
#'
#' \dontrun{
#' fars_map_state(3, 2014)   # error
#' }
#'
#' @export
fars_map_state <- function(state.num, year) {
  filename <- make_filename(year)
  data <- fars_read(filename)
  state.num <- as.integer(state.num)

  if(!(state.num %in% unique(data$STATE)))
    stop("invalid STATE number: ", state.num)
  data.sub <- dplyr::filter_(data, ~ STATE == state.num)
  if(nrow(data.sub) == 0L) {
    message("no accidents to plot")
    return(invisible(NULL))
  }
  is.na(data.sub$LONGITUD) <- data.sub$LONGITUD > 900
  is.na(data.sub$LATITUDE) <- data.sub$LATITUDE > 90
  with(data.sub, {
    maps::map("state", ylim = range(LATITUDE, na.rm = TRUE),
              xlim = range(LONGITUD, na.rm = TRUE))
    graphics::points(LONGITUD, LATITUDE, pch = 46)
  })
}