Imagine you are working with a dataset that contains information about sales transactions from a small retail shop. The data arrives in a messy form: some values are missing, column names are inconsistent, and there are duplicate rows. Your goal is to apply a complete data wrangling workflow using **Tidyverse** tools in R. You will need to import (simulate) the data, clean it by renaming columns and removing duplicates, transform it by creating new variables and handling missing values, and finally summarize the sales by product category to provide insights for the shop owner.

library(testthat)

source("user_code.R")
test_that("columns are renamed to tidy names", {
    expect_true(exists("clean_sales"), info = "'clean_sales' data frame must exist.")
    expected_names <- c("product_id", "product", "category", "sales_amount", "date_sold", "sales_amount_usd")
    expect_setequal(names(clean_sales), expected_names)
})

test_that("duplicate rows are removed", {
    # Duplicates: row 2 and 3, row 6 and 7 in messy_sales
    # Only one instance of each should remain after cleaning
    expect_true(exists("clean_sales"), info = "'clean_sales' data frame must exist.")
    expect_equal(nrow(clean_sales), 4, info = "There should be exactly 4 rows after removing duplicates and rows with missing date_sold.")
})

test_that("missing values in sales_amount are replaced with 0", {
    expect_true(exists("clean_sales"), info = "'clean_sales' data frame must exist.")
    expect_false(any(is.na(clean_sales$sales_amount)), info = "There should be no missing values in 'sales_amount'.")
    # The row with NA in 'sales_amount' should now be 0
    expect_true(any(clean_sales$sales_amount == 0), info = "Missing 'sales_amount' values must be replaced with 0.")
})

test_that("rows with missing date_sold are removed", {
    expect_true(exists("clean_sales"), info = "'clean_sales' data frame must exist.")
    expect_false(any(is.na(clean_sales$date_sold)), info = "There should be no missing values in 'date_sold'.")
})

test_that("sales_amount_usd column is correct", {
    expect_true(exists("clean_sales"), info = "'clean_sales' data frame must exist.")
    expect_true("sales_amount_usd" %in% names(clean_sales), info = "'sales_amount_usd' column must exist.")
    expect_equal(clean_sales$sales_amount_usd, clean_sales$sales_amount * 2, info = "'sales_amount_usd' must be twice 'sales_amount'.")
})

test_that("summary_sales shows total sales in USD for each category", {
    expect_true(exists("summary_sales"), info = "'summary_sales' data frame must exist.")
    expect_true(all(c("category", "total_sales") %in% names(summary_sales)), info = "'summary_sales' must have 'category' and 'total_sales' columns.")
    # Check values
    # After cleaning:
    # Fruit: Apple (10*2=20), Banana (15*2=30), Date (8*2=16) => 20+30+16=66
    # Vegetable: Carrot (0*2=0)
    fruit_sales <- summary_sales$total_sales[summary_sales$category == "Fruit"]
    veg_sales <- summary_sales$total_sales[summary_sales$category == "Vegetable"]
    expect_equal(fruit_sales, 66, info = "Total sales for 'Fruit' category should be 66 USD.")
    expect_equal(veg_sales, 0, info = "Total sales for 'Vegetable' category should be 0 USD.")
})


test_main.R

Master the essential data wrangling skills in R using the Tidyverse suite of packages. This course covers importing, cleaning, transforming, and summarizing data, empowering you to prepare datasets for analysis and visualization.

Comprehensive coverage of data wrangling techniques using Tidyverse in R.

Challenge: Data Wrangling Workflow

Lösung