A

Load necessary libraries

library(ggplot2) library(readr) library(forecast)

Load the data

data <- read.csv(“~/TCD Study/Data Mining/ShampooSales.csv”)

Inspect the first few rows of the data

print(head(data))

Check the column names to ensure they are correct

print(names(data))

Correct the date conversion to handle “Jan-95” format

data\(Month <- as.Date(paste0("01-", data\)Month), format=“%d-%b-%y”)

Check for any missing or infinite values

missing_month <- which(is.na(data\(Month)) missing_sales <- which(is.na(data\)Shampoo.Sales)) infinite_sales <- which(!is.finite(data$Shampoo.Sales))

Print rows with missing or infinite values

if (length(missing_month) > 0) print(data[missing_month, ]) if (length(missing_sales) > 0) print(data[missing_sales, ]) if (length(infinite_sales) > 0) print(data[infinite_sales, ])

Clean the data by removing rows with missing or infinite values

data_clean <- data[complete.cases(data) & is.finite(data$Shampoo.Sales), ]

Verify the cleaned data

print(head(data_clean))

Create the time plot

ggplot(data_clean, aes(x = Month, y = Shampoo.Sales)) + geom_line(color = “blue”) + labs( title = “Shampoo Sales Over Time”, x = “Month”, y = “Shampoo Sales” ) + theme_minimal() + theme( plot.title = element_text(hjust = 0.5, face = “bold”) )

B

Convert the data to a time series object

shampoo_ts <- ts(data_clean$Shampoo.Sales, frequency = 12, start = c(1980, 1))

Decompose the time series

shampoo_decomp <- decompose(shampoo_ts)

Plot the decomposed time series

autoplot(shampoo_decomp) + labs( title = “Decomposition of Shampoo Sales Time Series”, x = “Year”, y = “Shampoo Sales” ) + theme_minimal() + theme( plot.title = element_text(hjust = 0.5, face = “bold”) )