plot4.R

plot4 <- function () {
    # load the data for 1-2 Feb 2007
    pc1 <- loadPCdays()  
    
    # create the plot on a transparant background (is in the example provided)
    par (bg='transparent', mfrow= c(2,2), cex=0.6, omi=c(0, 0.05, 0, 0))
    
    # plot 1
    plot (pc1$DateTime, pc1$Global_active_power, type='l', ylab='Global Active Power', xlab='')
    
    #plot 2
    plot (pc1$DateTime, pc1$Voltage, type='l', ylab='Voltage', xlab='datetime')
    
    
    #plot 3
    plot (pc1$DateTime, pc1$Sub_metering_1, type= 'n', ylab='Energy sub metering', xlab='')
    points (pc1$DateTime, pc1$Sub_metering_1, type = 'l', col='black' )
    points (pc1$DateTime, pc1$Sub_metering_2, type = 'l', col='red' )
    points (pc1$DateTime, pc1$Sub_metering_3, type = 'l', col='blue' )
    # set the legend with 'text.with' to get the full text within the legend 
    legend ('topright', lty=1, col = c('black', 'red', 'blue'), 
            legend= c('Sub_metering_1','Sub_metering_2','Sub_metering_3' ),
            text.width = 72000, y.intersp=1.5, bty='n')
   
    #plot 4   
    plot (pc1$DateTime, pc1$Global_reactive_power, type='l', 
          ylab='Global_reactive_power', xlab='datetime')
    
    # copy to the png file with the same size as the example provided
    dev.copy(png, file = "plot4.png")
    dev.off()
    # reset the number of rows and columns
    par (mfrow= c(1,1))
}

loadPCdays <- function () {
    pc0 <<- read.csv('household_power_consumption.txt', sep=';', na.strings='?')
    # select the rows for Feb 1-2, 2007
    pc1 <- pc0[(pc0$Date=='1/2/2007')|(pc0$Date=='2/2/2007'),]
    #add DateTime column
    pc1$DateTime = strptime(paste (pc1$Date,pc1$Time, sep=' '), "%d/%m/%Y %H:%M:%S", tz='')
    pc1
}