plot6.R.txt

library(data.table)
library(ggplot2)

#LoadData : Load NEI and SCC data into global variables if not already present.
#           These variables are then accessed by plot functions.
#           NEI and SCC are loaded as data.table class.
#           If data files do not exist locally then they are downloaded
#           and unzipped. 
LoadData <- function() {
    
    if (!file.exists("summarySCC_PM25.rds") ||
            !file.exists("Source_Classification_Code.rds")) {
        
        if (!file.exists("FNEI_data.zip")) {
            print("Downloading FNEI_data.zip ...")
            download.file("https://d396qusza40orc.cloudfront.net/exdata%2Fdata%2FNEI_data.zip", 
                          "FNEI_data.zip", mode="wb")
            unzip("FNEI_data.zip")
        }
    }
    
    #Note: NEI and SCC are set in global environment
    if (!exists("NEI")) {
        print("Loading NEI data ...")
        NEI <<- as.data.table(readRDS("summarySCC_PM25.rds"))
    }
    if (!exists("SCC")) {
        print("Loading SCC data ...")
        SCC <<- as.data.table(readRDS("Source_Classification_Code.rds"))
    }
}


#plot6:     Plots graphs showing PM 2.5 comparisions of
#           emissions from motor vehicles between 
#           Baltimore City and LA  in USA
plot6 <- function () {
    
    LoadData() #NEI & SCC are now in global environment
    
    # Links:
    #   http://www.epa.gov/air/emissions/basic.htm
    #   http://www.epa.gov/air/emissions/index.htm
    #   http://www.epa.gov/cgi-bin/broker?polchoice=PM&_debug=0&_service=data&_program=dataprog.national_1.sas
    # Studying the links above it is understood that we need to grep
    # "Mobile Sources" from SCC level one and ON-ROAD point types from NEI
    
    scc.mobile <- grepl("mobile", SCC$SCC.Level.One, ignore.case=TRUE)
    scc <- SCC[scc.mobile,]$SCC
    
    #subset data of motor vehicles in Baltimore
    nei <- NEI[(fips == "24510") & (SCC %in% scc) & (type=="ON-ROAD"), 
               sum(Emissions), by=year]
    setnames(nei, 2, "balt")
    
    #subset data of motor vehicles in LA
    nei.la <- NEI[(fips == "06037") & (SCC %in% scc) & (type=="ON-ROAD"), 
                  sum(Emissions), by=year]
    setnames(nei.la, 2, "la")
    nei[, la := nei.la$la]
    
    #add relative (in %) and absolute differences
    nei[, c("balt.per","la.per", "balt.abs", "la.abs") := 
            list(balt/max(balt), la/max(la), 
                 balt - head(balt, n=1), la - head(la, n=1))]
    
    #plot 3 graphs showing actual values, absolute differences and relative differences
    par(mfrow = c(1,3), oma = c(0, 0, 3, 0))
    
    plot(nei$year, nei$la, ylim=c(0, 4700), 
         type="n", xaxt="n",
         xlab="Year", ylab="Emissions (in tons)")
    lines(nei$year, nei$balt, col="red")
    lines(nei$year, nei$la, col="blue")
    axis(1, at=nei$year, labels=nei$year)
    legend("topleft", lty=c(1,1), col=c("red", "blue"),
           legend=c("Baltimore", "LA"))
    
    plot(nei$year, nei$balt.abs, ylim=c(-300, 700), 
         type="n", xaxt="n",
         xlab="Year", ylab="Absolute differeces (in tons)")
    lines(nei$year, nei$balt.abs, col="red")
    lines(nei$year, nei$la.abs, col="blue")
    axis(1, at=nei$year, labels=nei$year)
    legend("topleft", lty=c(1,1), col=c("red", "blue"), 
           legend=c("Baltimore", "LA"))
    
    plot(nei$year, nei$balt.per, 
         type="n", xaxt="n",
         xlab="Year", ylab="Relative %age differeces")
    lines(nei$year, nei$balt.per, col="red")
    lines(nei$year, nei$la.per, col="blue")
    axis(1, at=nei$year, labels=nei$year)
    legend("bottomleft", lty=c(1,1), col=c("red", "blue"),
           legend=c("Baltimore", "LA"))
    
    title("top", outer=TRUE, cex.main=2.0, 
          main="Motor vehicle PM 2.5 emissions comparision of Baltimore and LA")
    
    #save the data to png
    dev.copy(png, file="plot6.png", width=1000, height=500)
    dev.off()
    
    #reset par to original values
    par(mfrow = c(1,1), oma = c(0,0,0,0))
    
    #return the data to the caller
    nei
}

plot6()