# Load the tidyverse library
library(tidyverse)
# Read in the car sales data
# Make sure the data file is in your lab_0/data/ folder
car_data <- read_csv("../data/car_sales_data.csv")Lab 0: Getting Started with dplyr
Your First Data Analysis
Overview
Welcome to your first lab! In this (not graded) assignment, you’ll practice the fundamental dplyr operations I overviewed in class using car sales data. This lab will help you get comfortable with:
- Basic data exploration
- Column selection and manipulation
- Creating new variables
- Filtering data
- Grouping and summarizing
Instructions: Copy this template into your portfolio repository under a lab_0/ folder, then complete each section with your code and answers. You will write the code under the comment section in each chunk. Be sure to also copy the data folder into your lab_0 folder.
Setup
Exercise 1: Getting to Know Your Data
1.1 Data Structure Exploration
Explore the structure of your data and answer these questions:
# Use glimpse() to see the data structure
# Check the column names
# Look at the first few rows
head(car_data)# A tibble: 6 × 7
Manufacturer Model `Engine size` `Fuel type` `Year of manufacture` Mileage
<chr> <chr> <dbl> <chr> <dbl> <dbl>
1 Ford Fiesta 1 Petrol 2002 127300
2 Porsche 718 Caym… 4 Petrol 2016 57850
3 Ford Mondeo 1.6 Diesel 2014 39190
4 Toyota RAV4 1.8 Hybrid 1988 210814
5 VW Polo 1 Petrol 2006 127869
6 Ford Focus 1.4 Petrol 2018 33603
# ℹ 1 more variable: Price <dbl>Questions to answer: - How many rows and columns does the dataset have? - What types of variables do you see (numeric, character, etc.)? - Are there any column names that might cause problems? Why?
Your answers: - Rows: 50000 - Columns: 7
- Variable types: Strings and Numbers - Problematic names: Year of manufacture is stored as a number rather than in a date format (date would be easier to do time based calculations on). Also, the column name is way too long.
1.2 Tibble vs Data Frame
Compare how tibbles and data frames display:
# Look at the tibble version (what we have)
car_data# A tibble: 50,000 × 7
Manufacturer Model `Engine size` `Fuel type` `Year of manufacture` Mileage
<chr> <chr> <dbl> <chr> <dbl> <dbl>
1 Ford Fiesta 1 Petrol 2002 127300
2 Porsche 718 Cay… 4 Petrol 2016 57850
3 Ford Mondeo 1.6 Diesel 2014 39190
4 Toyota RAV4 1.8 Hybrid 1988 210814
5 VW Polo 1 Petrol 2006 127869
6 Ford Focus 1.4 Petrol 2018 33603
7 Ford Mondeo 1.8 Diesel 2010 86686
8 Toyota Prius 1.4 Hybrid 2015 30663
9 VW Polo 1.2 Petrol 2012 73470
10 Ford Focus 2 Diesel 1992 262514
# ℹ 49,990 more rows
# ℹ 1 more variable: Price <dbl># Convert to regular data frame and display
car_df <- as.data.frame(car_data)
head(car_df) Manufacturer Model Engine size Fuel type Year of manufacture Mileage
1 Ford Fiesta 1.0 Petrol 2002 127300
2 Porsche 718 Cayman 4.0 Petrol 2016 57850
3 Ford Mondeo 1.6 Diesel 2014 39190
4 Toyota RAV4 1.8 Hybrid 1988 210814
5 VW Polo 1.0 Petrol 2006 127869
6 Ford Focus 1.4 Petrol 2018 33603
Price
1 3074
2 49704
3 24072
4 1705
5 4101
6 29204Question: What differences do you notice in how they print?
Your answer: In the rendered q markdown view, the tibble is much cleaner. It prints the size of the tibble, but only shows 10 rows and however many columns fit properly in the view. The dataframe prints the all 5000 rows! I’ve edited the code to only print the head() to avoid this.
Exercise 2: Basic Column Operations
2.1 Selecting Columns
Practice selecting different combinations of columns:
# Select just Model and Mileage columns
model_mileage <- car_data %>%
select(Model, Mileage)
# Select Manufacturer, Price, and Fuel type
manufacturer_price_fuel <- car_data %>%
select(Manufacturer, Price, `Fuel type`)
# Challenge: Select all columns EXCEPT Engine Size
car_data_sans_engine <- car_data %>%
select(!`Engine size`)2.2 Renaming Columns
Let’s fix a problematic column name:
# Rename 'Year of manufacture' to year
car_data <- car_data %>%
rename(year = `Year of manufacture`)
# Check that it worked
names(car_data)[1] "Manufacturer" "Model" "Engine size" "Fuel type" "year"
[6] "Mileage" "Price" Question: Why did we need back-ticks around Year of manufacture but not around year?
Your answer: There are spaces in the column name, so the backticks help R understand what the full name.
Exercise 3: Creating New Columns
3.1 Calculate Car Age
# Create an 'age' column (2025 minus year of manufacture)
car_data <- car_data %>%
mutate(age = 2025-year)
# Create a mileage_per_year column
car_data <- car_data %>%
mutate(mileage_per_year = Mileage/age)
# Look at your new columns
#select(car_data, Model, year, age, Mileage, mileage_per_year)
car_data %>%
select(Model, year, age, Mileage, mileage_per_year)# A tibble: 50,000 × 5
Model year age Mileage mileage_per_year
<chr> <dbl> <dbl> <dbl> <dbl>
1 Fiesta 2002 23 127300 5535.
2 718 Cayman 2016 9 57850 6428.
3 Mondeo 2014 11 39190 3563.
4 RAV4 1988 37 210814 5698.
5 Polo 2006 19 127869 6730.
6 Focus 2018 7 33603 4800.
7 Mondeo 2010 15 86686 5779.
8 Prius 2015 10 30663 3066.
9 Polo 2012 13 73470 5652.
10 Focus 1992 33 262514 7955.
# ℹ 49,990 more rows3.2 Categorize Cars
# Create a price_category column where if price is < 15000, it is coded as budget, between 15000 and 30000 is midrange and greater than 30000 is luxury (use case_when)
car_data <- car_data %>%
mutate(price_category = case_when(
Price > 30000 ~ "luxury",
Price >= 15000 & Price <= 30000 ~ "midrange",
.default = "budget"
))
# Check your categories select the new column and show it
car_data %>%
select(Manufacturer, Model, price_category)# A tibble: 50,000 × 3
Manufacturer Model price_category
<chr> <chr> <chr>
1 Ford Fiesta budget
2 Porsche 718 Cayman luxury
3 Ford Mondeo midrange
4 Toyota RAV4 budget
5 VW Polo budget
6 Ford Focus midrange
7 Ford Mondeo budget
8 Toyota Prius luxury
9 VW Polo budget
10 Ford Focus budget
# ℹ 49,990 more rowsExercise 4: Filtering Practice
4.1 Basic Filtering
# Find all Toyota cars
car_data %>%
filter(Manufacturer == "Toyota")# A tibble: 12,554 × 10
Manufacturer Model `Engine size` `Fuel type` year Mileage Price age
<chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 Toyota RAV4 1.8 Hybrid 1988 210814 1705 37
2 Toyota Prius 1.4 Hybrid 2015 30663 30297 10
3 Toyota RAV4 2.2 Petrol 2007 79393 16026 18
4 Toyota Yaris 1.4 Petrol 1998 97286 4046 27
5 Toyota RAV4 2.4 Hybrid 2003 117425 11667 22
6 Toyota Yaris 1.2 Petrol 1992 245990 720 33
7 Toyota RAV4 2 Hybrid 2018 28381 52671 7
8 Toyota Prius 1 Hybrid 2003 115291 6512 22
9 Toyota Prius 1 Hybrid 1990 238571 961 35
10 Toyota Prius 1.8 Hybrid 2017 31958 38961 8
# ℹ 12,544 more rows
# ℹ 2 more variables: mileage_per_year <dbl>, price_category <chr># Find cars with mileage less than 30,000
car_data %>%
filter(Mileage < 30000)# A tibble: 5,402 × 10
Manufacturer Model `Engine size` `Fuel type` year Mileage Price age
<chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 Toyota RAV4 2 Hybrid 2018 28381 52671 7
2 VW Golf 2 Petrol 2020 18985 36387 5
3 BMW M5 4 Petrol 2017 22759 97758 8
4 Toyota RAV4 2.4 Petrol 2018 24588 49125 7
5 VW Golf 2 Hybrid 2018 25017 36957 7
6 Porsche 718 Cayman 2.4 Petrol 2021 14070 69526 4
7 Ford Focus 1.8 Petrol 2020 22371 40336 5
8 Ford Mondeo 1.6 Diesel 2015 21834 28435 10
9 VW Passat 1.6 Diesel 2018 22122 36634 7
10 VW Passat 1.4 Diesel 2020 21413 39310 5
# ℹ 5,392 more rows
# ℹ 2 more variables: mileage_per_year <dbl>, price_category <chr># Find luxury cars (from price category) with low mileage
car_data %>%
filter(Mileage < 30000 & price_category == "luxury")# A tibble: 3,256 × 10
Manufacturer Model `Engine size` `Fuel type` year Mileage Price age
<chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 Toyota RAV4 2 Hybrid 2018 28381 52671 7
2 VW Golf 2 Petrol 2020 18985 36387 5
3 BMW M5 4 Petrol 2017 22759 97758 8
4 Toyota RAV4 2.4 Petrol 2018 24588 49125 7
5 VW Golf 2 Hybrid 2018 25017 36957 7
6 Porsche 718 Cayman 2.4 Petrol 2021 14070 69526 4
7 Ford Focus 1.8 Petrol 2020 22371 40336 5
8 VW Passat 1.6 Diesel 2018 22122 36634 7
9 VW Passat 1.4 Diesel 2020 21413 39310 5
10 Toyota RAV4 2.4 Petrol 2021 6829 66031 4
# ℹ 3,246 more rows
# ℹ 2 more variables: mileage_per_year <dbl>, price_category <chr>4.2 Multiple Conditions
# Find cars that are EITHER Honda OR Nissan
# Note: I used %in% to avoid writing Manufacturer twice.
car_data %>%
filter(Manufacturer %in% c("Honda", "Nissan"))# A tibble: 0 × 10
# ℹ 10 variables: Manufacturer <chr>, Model <chr>, Engine size <dbl>,
# Fuel type <chr>, year <dbl>, Mileage <dbl>, Price <dbl>, age <dbl>,
# mileage_per_year <dbl>, price_category <chr># Find cars with price between $20,000 and $35,000
# Note: I used between() to have a more concise query.
car_data %>%
filter(between(Price, 20000, 35000 ))# A tibble: 7,301 × 10
Manufacturer Model `Engine size` `Fuel type` year Mileage Price age
<chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 Ford Mondeo 1.6 Diesel 2014 39190 24072 11
2 Ford Focus 1.4 Petrol 2018 33603 29204 7
3 Toyota Prius 1.4 Hybrid 2015 30663 30297 10
4 Toyota Prius 1.4 Hybrid 2016 43893 29946 9
5 Toyota Prius 1.4 Hybrid 2016 43130 30085 9
6 VW Passat 1.6 Petrol 2016 64344 23641 9
7 Ford Mondeo 1.6 Diesel 2015 21834 28435 10
8 BMW M5 4.4 Petrol 2008 109941 31711 17
9 BMW Z4 2.2 Petrol 2014 61332 26084 11
10 Porsche 911 3.5 Petrol 2003 107705 24378 22
# ℹ 7,291 more rows
# ℹ 2 more variables: mileage_per_year <dbl>, price_category <chr># Find diesel cars less than 10 years old
car_data %>%
filter (age < 10 & `Fuel type` == "Diesel")# A tibble: 2,040 × 10
Manufacturer Model `Engine size` `Fuel type` year Mileage Price age
<chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 Ford Fiesta 1 Diesel 2017 38370 16257 8
2 VW Passat 1.6 Diesel 2018 22122 36634 7
3 VW Passat 1.4 Diesel 2020 21413 39310 5
4 BMW X3 2 Diesel 2018 27389 44018 7
5 Ford Mondeo 2 Diesel 2016 51724 28482 9
6 Porsche Cayenne 2.6 Diesel 2019 20147 76182 6
7 VW Polo 1.2 Diesel 2018 37411 19649 7
8 Ford Mondeo 1.8 Diesel 2016 29439 30886 9
9 Ford Mondeo 1.4 Diesel 2020 18929 37720 5
10 Ford Mondeo 1.4 Diesel 2018 42017 28904 7
# ℹ 2,030 more rows
# ℹ 2 more variables: mileage_per_year <dbl>, price_category <chr>Question: How many diesel cars are less than 10 years old?
Your answer: 2040
Exercise 5: Grouping and Summarizing
5.1 Basic Summaries
# Calculate average price by manufacturer
avg_price_by_brand <- car_data %>%
group_by(Manufacturer) %>%
summarize(avg_price = mean(Price, na.rm = TRUE))
avg_price_by_brand# A tibble: 5 × 2
Manufacturer avg_price
<chr> <dbl>
1 BMW 24429.
2 Ford 10672.
3 Porsche 29104.
4 Toyota 14340.
5 VW 10363.# Calculate average mileage by fuel type
avg_price_by_fuel_type <- car_data %>%
group_by(`Fuel type`) %>%
summarize(avg_price = mean(Price, na.rm = TRUE))
avg_price_by_fuel_type# A tibble: 3 × 2
`Fuel type` avg_price
<chr> <dbl>
1 Diesel 13145.
2 Hybrid 14949.
3 Petrol 13691.# Count cars by manufacturer
count_by_manufacturer <- car_data %>%
group_by(Manufacturer) %>%
summarize("Count" = n())
count_by_manufacturer# A tibble: 5 × 2
Manufacturer Count
<chr> <int>
1 BMW 4965
2 Ford 14959
3 Porsche 2609
4 Toyota 12554
5 VW 149135.2 Categorical Summaries
# Frequency table for price categoriesSubmission Notes
To submit this lab: 1. Make sure your code runs without errors 2. Fill in all the “[YOUR ANSWER]” sections and complete all of the empty code! 3. Save this file in your portfolio’s lab_0/ folder 4. Commit and push to GitHub 5. Check that it appears on your GitHub Pages portfolio site
Questions? Post on the canvas discussion board or come to office hours!