plot1.R

###
### this function plots an histogram of the global active power for the household power consumption data set
### The data set is downloaded and unzipped if this was not done before. 
###
plot1 <- function(){
    
    # define some constants 
    fileURL <- "https://d396qusza40orc.cloudfront.net/exdata%2Fdata%2Fhousehold_power_consumption.zip"
    zipName <- "exdata_data_household_power_consumption.zip"
    fileName <- "household_power_consumption.txt"
    dateFormat <- "%d/%m/%Y"
    timeFormat <- "%H:%M:%S"
    startDate <- "1/2/2007"
    stopDate <- "2/2/2007"
    plotFile <- "plot1.png"
    Sys.setlocale("LC_TIME", "English")
    
    # make sure the data set is available on disk
    if(!file.exists(fileName)){
        if (!file.exists(zipName)){
            download.file(fileURL, zipName, mode="wb")
        }
        unzip(zipName)
    }
    
    # create new types in order to parse the data directly during loading
    setAs("character","hpcDate", function(from) as.Date(from, format=dateFormat) )
    setClass('hpcDate')
    setAs("character","hpcTime", function(from) strptime(from, format=timeFormat) )
    setClass('hpcTime')
    setAs("character","hpcNumeric", function(from) suppressWarnings(as.numeric(from)) ) # numeric fails for "?" values
    setClass('hpcNumeric')
    
    # loading and general formatting the data set
    colClasses <- c(
        'hpcDate', # col 1 is the date
        'hpcTime', # col 2 is the time
        'hpcNumeric', # col 3 is the Global_active_power
        'hpcNumeric', # col 4 is the Global_reactive_power
        'hpcNumeric', # col 5 is the Voltage
        'hpcNumeric', # col 6 is the Global_intensity
        'hpcNumeric', # col 7 is the Sub_metering_1
        'hpcNumeric', # col 8 is the Sub_metering_2
        'hpcNumeric'  # col 9 is the Sub_metering_3
    );    
    raw <- read.table(fileName, sep=";", colClasses=colClasses, header=TRUE)
    data <- raw[raw$Date==as.Date(startDate, format=dateFormat) | raw$Date==as.Date(stopDate, format=dateFormat),]
    # we don't have to check for NA's because there are none in the range we're interested in, 
    # but I keep it for the sake of completeness
    data <- data[!is.na(data[,3]), ] # no need to check other columns. Columns 3 to 9 have a ? together.
    
    # plot the chart
    # the reference is 504X504 while the one I produce is 480X480
    png(plotFile, width = 480, height = 480)
    hist(data$Global_active_power, col="red", main="Global Active Power", xlab="Global Active Power (kilowatts)")
    dev.off()
    
    # return the data set as used for plotting
    data
}