Exploratory Data Analysis PA1

Purpose

The purpose of this assignment is to use UC Irvine’s Machine Learning Respository’s data to validate the household electric power consumptions, please refer to this particular electrical power consumption data link

Step Wise Implementation

• 1. Load the data.table package to execute while download & unzip the dataset from UCI repo by using download.file and unzip command for exporting the data file into the working directory.
     
• 2. Use data.table’s fread function to read the datafile while identify the na.strings (in this case as ? mark for the NA values) to import into the desired data table format.
     
• 3. Use str to examine the current variable’s characteristics, and notice the date and time are separated as char.
     
• 4. Re-arrange the data column’s format to desired format by either using as.Date or as.POSIXct to transform the data to ensure the data format is within the designated part.
     
• 5. Filter the necessary time frame so the data is with in “2007-02-01” to “2007-02-02”
     
• 6. Setup the respective plot dimension and layout before implementing the plotting command for designated pictures.
     
• 7. Plot respective plot by either using single command to reach required parts or use multiple lines.
     
     

Code Chunks

# Plot 1 Setup
# Load the data.table package for the function execution
library(data.table)

# Define URL to download and unzip the text file.
url <- "https://d396qusza40orc.cloudfront.net/exdata%2Fdata%2Fhousehold_power_consumption.zip"
download.file(url,"PA01.zip")
unzip('PA01.zip')

# Export data table for power consumption from extract file.
PDB <- data.table::fread(input = "household_power_consumption.txt"
                         , na.strings="?"
)

# Check the current structure of the exported data table to identify the variable characteristics
str(PDB)

# Convert the active power as a numeric value to prevent scientific notation.
PDB[, Global_active_power := lapply(.SD, as.numeric), .SDcols = c("Global_active_power")]

# Change Date Column from char to date data type.
PDB[, Date := lapply(.SD, as.Date, "%d/%m/%Y"), .SDcols = c("Date")]

# Filter Dates for 2007-02-01 and 2007-02-02
PDB_F1 <- PDB[(Date >= "2007-02-01") & (Date <= "2007-02-02")]


# Setup plot dimension format
png("plot1.png", width=480, height=480)

# Plot 1
hist(PDB_F1[, Global_active_power], main="Global Active Power", 
     xlab="Global Active Power (kilowatts)", ylab="Frequency",col="Red")

dev.off()

# Plot 2 Setup
# Load the data.table package for the function execution
library(data.table)

# Define URL to download and unzip the text file.
url <- "https://d396qusza40orc.cloudfront.net/exdata%2Fdata%2Fhousehold_power_consumption.zip"
download.file(url,"PA01.zip")
unzip('PA01.zip')

# Export data table for power consumption from extract file.
PDB <- data.table::fread(input = "household_power_consumption.txt"
                         , na.strings="?"
)

# Check the current structure of the exported data table to identify the variable characteristics
str(PDB)

# Convert the active power as a numeric value
PDB[, Global_active_power := lapply(.SD, as.numeric), .SDcols = c("Global_active_power")]

# Add the time with date as a POSIXct object for timeline filtering
PDB[, dateTime := as.POSIXct(paste(Date, Time), format = "%d/%m/%Y %H:%M:%S")]

# Filter Dates for 2007-02-01 and 2007-02-02
PDB_F2 <- PDB[(dateTime >= "2007-02-01") & (dateTime < "2007-02-03")]


# Setup plot dimension format
png("plot2.png", width=480, height=480)

# Plot 2
plot(x = PDB_F2[,dateTime], y = PDB_F2[,Global_active_power], type = "l"
     ,xlab="",ylab="Global Active Power (kilowatts)")

dev.off()

# Plot 3 Setup
# Load the data.table package for the function execution
library(data.table)

# Define URL to download and unzip the text file.
url <- "https://d396qusza40orc.cloudfront.net/exdata%2Fdata%2Fhousehold_power_consumption.zip"
download.file(url,"PA01.zip")
unzip('PA01.zip')

# Export data table for power consumption from extract file.
PDB <- data.table::fread(input = "household_power_consumption.txt"
                         , na.strings="?"
)

# Check the current structure of the exported data table to identify the variable characteristics
str(PDB)

# Convert the active power as a numeric value
PDB[, Global_active_power := lapply(.SD, as.numeric), .SDcols = c("Global_active_power")]

# Add the time with date as a POSIXct object for timeline filtering
PDB[, dateTime := as.POSIXct(paste(Date, Time), format = "%d/%m/%Y %H:%M:%S")]

# Filter Dates for 2007-02-01 and 2007-02-02
PDB_F3 <- PDB[(dateTime >= "2007-02-01") & (dateTime < "2007-02-03")]


# Setup plot dimension format
png("plot3.png", width=480, height=480)

# Plot 3's core structure of 1st sub metering
plot(x = PDB_F3[,dateTime], y = PDB_F3[,Sub_metering_1], type = "l"
     ,xlab="",ylab="Energy sub metering")

# Use Lines command to add the remaining sub meters
lines(x = PDB_F3[,dateTime], y = PDB_F3[,Sub_metering_2],col="red")
lines(x = PDB_F3[,dateTime], y = PDB_F3[,Sub_metering_3],col="blue")
legend("topright", col = c("black","red","blue")
       ,c("Sub_metering_1 ","Sub_metering_2 ","Sub_metering_3 ")
       , lty = c(1,1), lwd=c(1,1))

dev.off()

# Plot 4 Setup
# Load the data.table package for the function execution
library(data.table)

# Define URL to download and unzip the text file.
url <- "https://d396qusza40orc.cloudfront.net/exdata%2Fdata%2Fhousehold_power_consumption.zip"
download.file(url,"PA01.zip")
unzip('PA01.zip')

# Export data table for power consumption from extract file.
PDB <- data.table::fread(input = "household_power_consumption.txt"
                         , na.strings="?"
)

# Check the current structure of the exported data table to identify the variable characteristics
str(PDB)

# Convert the active power as a numeric value
PDB[, Global_active_power := lapply(.SD, as.numeric), .SDcols = c("Global_active_power")]

# Add the time with date as a POSIXct object for timeline filtering
PDB[, dateTime := as.POSIXct(paste(Date, Time), format = "%d/%m/%Y %H:%M:%S")]

# Filter Dates for 2007-02-01 and 2007-02-02
PDB_F4 <- PDB[(dateTime >= "2007-02-01") & (dateTime < "2007-02-03")]


# Setup plot dimension format
png("plot4.png", width=480, height=480)

# Setup the 2x2 matrix form for the plog grid
par(mfrow = c(2,2))

# 1st plot (coming from plot 2)
plot(x = PDB_F4[,dateTime], y = PDB_F4[,Global_active_power], type = "l"
     ,xlab="",ylab="Global Active Power (kilowatts)")

# 2nd plot 
plot(x = PDB_F4[,dateTime], y = PDB_F4[,Voltage], type = "l"
     ,xlab="datetime",ylab="Voltage")

# 3rd plot (coming from plot 3)
plot(x = PDB_F4[,dateTime], y = PDB_F4[,Sub_metering_1], type = "l"
     ,xlab="",ylab="Energy sub metering")
lines(x = PDB_F4[,dateTime], y = PDB_F4[,Sub_metering_2],col="red")
lines(x = PDB_F4[,dateTime], y = PDB_F4[,Sub_metering_3],col="blue")
legend("topright", col = c("black","red","blue")
       ,c("Sub_metering_1 ","Sub_metering_2 ","Sub_metering_3 ")
       , lty = c(1,1), bty="n", cex = .5)

# 4th plot
plot(x = PDB_F4[,dateTime], y = PDB_F4[,Global_reactive_power], type = "l"
     ,xlab="datetime",ylab="Global_reactive_power")

dev.off()