R/fars_functions.R

In EnriquePH/FARS: Read Fatality Analysis Reporting System (FARS) data

#  ----------------------------------------------------------------------------
#  COURSERA: BUILDING R PACKAGES
#  Peer-graded Assignment
#  File: fars_functions.R
#  (c) 2017 - Enrique Pérez Herrero
#  email: eph.project1500@gmail.com
#  GNU GENERAL PUBLIC LICENSE Version 3, 29 June 2007
#  Start: 08/May/2017
#  End:   18/May/2017
#  ---------------------------------------------------------------------------

#' Read file with FARS data
#'
#' This function reads data from .csv file, stored on disk, from the \strong{US
#' National Highway Traffic Safety Administration's} \emph{Fatality Analysis
#' Reporting System} (FARS), which is a nationwide census, providing the
#' American public yearly data, regarding fatal injuries suffered in motor
#' vehicle traffic crashes.
#'
#' @details For more information, see:
#' \itemize{
#'   \item{\url{https://www.nhtsa.gov/research-data/fatality-analysis-reporting-system-fars}}
#'   \item{\url{https://en.wikipedia.org/wiki/Fatality_Analysis_Reporting_System}}
#' }
#' @importFrom readr read_csv
#' @importFrom dplyr tbl_df
#'
#' @param filename A character string with the name of the file to read, see
#'   notes.
#'
#' @return A data frame with data readed from the csv file, or an error if the
#'   file does not exists.
#'
#' @examples
#' library(dplyr)
#' library(readr)
#' yr <- 2015
#' data <- yr %>%
#'   make_filename %>%
#'   fars_read
#' head(data)
#' @note To generate file name use: \code{\link{make_filename}}
#' @seealso \link{make_filename}
#' @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 data file name
#'
#' Make .csv data file name related to the given \code{year}
#' The function does not check if the file is available.
#'
#' @param year A string or an integer with the input \code{year}
#'
#' @return This function returns a string with the data file name for a given
#'   year, and the file path within the package.
#'
#' @examples
#' make_filename(2013)
#' @seealso \link{fars_read}
#' @export
make_filename <- function(year) {
        year <- as.integer(year)
        system.file("extdata",
                    sprintf("accident_%d.csv.bz2", year),
                    package = "fars",
                    mustWork = TRUE)
}

#' Read FARS years
#'
#' Ancillary function used by \code{fars_summarize_years}
#' @param years A vector with a list of years
#'
#' @importFrom dplyr mutate_
#' @importFrom dplyr select_
#' @importFrom magrittr "%>%"
#
#' @return A data.frame including entries in data by month, or NULL if the
#'  \code{year} is not valid
#'
#' @seealso \link{fars_read}
#' @seealso \link{make_filename}
#' @seealso \link{fars_summarize_years}
#' @examples
#' fars_read_years(2013)
#' @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)
                })
        })
}

#' Summarize FARS data by years
#'
#' This function summarizes yearly accidents data, by month
#' @param years A vector with a list of years to summarize by.
#'
#' @return A data.frame with number of accidents by years summarized by month
#' @importFrom dplyr bind_rows
#' @importFrom dplyr group_by_
#' @importFrom dplyr summarize_
#' @importFrom tidyr spread_
#' @importFrom magrittr "%>%"
#' @seealso \link{fars_read_years}
#' @examples
#' plot(fars_summarize_years(2015))
#' fars_summarize_years(c(2015, 2014))
#' @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")
}

#' Display accidents map by state and year
#'
#' Displays a plot with a state map including the accidents location by year
#' If the \code{state.num} is invalid the function shows an error
#' @param state.num An Integer with the State Code
#' \tabular{cc}{
#'   \strong{State Code} \tab \strong{State Name}    \cr
#'   01 \tab  Alabama              \cr
#'   02 \tab  Alaska               \cr
#'   04 \tab  Arizona              \cr
#'   05 \tab  Arkansas             \cr
#'   06 \tab  California           \cr
#'   08 \tab  Colorado             \cr
#'   09 \tab  Connecticut          \cr
#'   10 \tab  Delaware             \cr
#'   11 \tab  District of Columbia \cr
#'   12 \tab  Florida              \cr
#'   13 \tab  Georgia              \cr
#'   15 \tab  Hawaii               \cr
#'   16 \tab  Idaho                \cr
#'   17 \tab  Illinois             \cr
#'   18 \tab  Indiana              \cr
#'   19 \tab  Iowa                 \cr
#'   20 \tab  Kansas               \cr
#'   21 \tab  Kentucky             \cr
#'   22 \tab  Louisiana            \cr
#'   23 \tab  Maine                \cr
#'   24 \tab  Maryland             \cr
#'   25 \tab  Massachusetts        \cr
#'   26 \tab  Michigan             \cr
#'   27 \tab  Minnesota            \cr
#'   28 \tab  Mississippi          \cr
#'   29 \tab  Missouri             \cr
#'   30 \tab  Montana              \cr
#'   31 \tab  Nebraska             \cr
#'   32 \tab  Nevada               \cr
#'   33 \tab  New Hampshire        \cr
#'   34 \tab  New Jersey           \cr
#'   35 \tab  New Mexico           \cr
#'   36 \tab  New York             \cr
#'   37 \tab  North Carolina       \cr
#'   38 \tab  North Dakota         \cr
#'   39 \tab  Ohio                 \cr
#'   40 \tab  Oklahoma             \cr
#'   41 \tab  Oregon               \cr
#'   42 \tab  Pennsylvania         \cr
#'   43 \tab  Puerto Rico          \cr
#'   44 \tab  Rhode Island         \cr
#'   45 \tab  South Carolina       \cr
#'   46 \tab  South Dakota         \cr
#'   47 \tab  Tennessee            \cr
#'   48 \tab  Texas                \cr
#'   49 \tab  Utah                 \cr
#'   50 \tab  Vermont              \cr
#'   51 \tab  Virginia             \cr
#'   52 \tab  Virgin Islands       \cr
#'   53 \tab  Washington           \cr
#'   54 \tab  West Virginia        \cr
#'   55 \tab  Wisconsin            \cr
#'   56 \tab  Wyoming
#' }
#' @param year A string, or an integer, with the input \code{year}
#'
#' @importFrom maps map
#' @importFrom dplyr filter_
#' @importFrom graphics points
#' @return None
#' @seealso \link{fars_read}
#' @seealso \link{make_filename}
#' @references 2014 FARS/NASS GES Coding and Validation Manual
#' @examples
#' \dontrun{
#' fars_map_state(49, 2015)
#' }
#' @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$STATE))) {
    stop("invalid STATE number: ", state.num)
  }
  data.sub <- dplyr::filter_(data, .dots = paste0("STATE==", state.num))
  if(nrow(data.sub) == 0L) {
    message("no accidents to plot")
    return(invisible(NULL))
  }
  is.na(data.sub$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)
  })
}