Creating new columns with dplyr

> Combining tables in dplyr

# Create a time_on_market column using the difference of today’s year and the year_listed

airbnb_listings %>%
x1 x2 x1 x2

Data Manipulation with dplyr in R mutate(time_on_market = 2022 - year_listed)

1
3
2
6
1
4
2
6

Cheat Sheet
# Create a full_address column by combining city and country
5 4 2 5
airbnb_listings %>%

transmute(full_address = paste(city, country))

df_1 df_2

Learn R online at www.DataCamp.com # Add the number of observations for a column (e.g., number of listings per city)
# Appending a table to the right side (horizontal) of another

bind_cols(df_1, df_2)
airbnb_listings %>%

add_count(city)
# Appending a table to the bottom (vertical) of another

bind_rows(df_1, df_2)

> Helpful Syntax

Working with rows # Combining rows that exist in both tables and dropping duplicates

union(df_1, df_2)
Installing and loading dplyr
# Filter rows on one condition (e.g., country)
# Finding identical columns in both tables

# Install dplyr through tidyverse

airbnb_listings %>%
intersect(df_1, df_2)
install.packages(“tidyverse”)

filter(country=="France")
# Finding rows that don’t exist in another table

# Install it directly
# Filter OR more conditions (country OR number_of_rooms)

on two setdiff(df_1, df_2)

install.packages(“dplyr”)

airbnb_listings %>%

filter(country=="France" | number_of_rooms > 3)

# Load dplyr into R

library(dplyr) # Filter AND more conditions (country AND

on two number_of_rooms)

The %>% operator

airbnb_listings %>%

filter(country=="France" & number_of_rooms > 3) > Joining Tables with dplyr

%>% is a special operator in R found in the magrittr and dplyr packages. %>% lets you pass objects to functions # Filter by checking if a value exists in another set of values

elegantly, and helps you make your code more readable. Consider this example of choosing columns a and b from the airbnb_listings %>%
To showcase joins in dplyr, we’ll use an additional dataset containing details on host_listings for airbnb listings
dataframe df filter(country %in% c("Japan", "France"))
# Without the %>% operator
airbnb_listings
# Filter rows based on index of rows (e.g., first 3 rows)
listing_id city country number_of_rooms year_listed
select(df, a, b)

Airbnb_listings %>%
1 Paris France 5 2018
slice(1:3) 2 Tokyo Japan 2 2017
# By using the %>% operator

df %>% select(a, b) 3 New York USA 2 2022

# Sort rows by values in a column in ascending order

airbnb_listings %>%

host_listings
arrange(number_of_rooms)
host_id name listing_id number_of_reviews

> Dataset used throughout this cheat sheet # Sort rows by values in a column in descending order

airbnb_listings %>%

1
2
Jen Bricker
Richie Cotton
1
2
34
12
3 Raven Todd Dasliva 3 55
arrange(desc(city))
Throughout this cheat sheet, we weill be using this example dataset called airbnb_listings, containing Airbnb
listings with data on their location, year listed, number of rooms, and more. # Remove duplicate rows in all the dataset

airbnb_listings
airbnb_listings %>%

distinct() Joining tables in dplyr

listing_id city country number_of_rooms year_listed
1 Paris France 5 2018 # Find unique values in the country column

airbnb_listings %>%
Inner Join
2 Tokyo Japan 2 2017
3 New York USA 2 2022 distinct(country) # Returns only records where a joining field finds a match in both tables.

airbnb_listings %>%

# Select rows based on top-n values of a column (e.g., top 3 listings with the highest amount inner_join(host_listings, by="listing_id")
of rooms)

> Transforming data with dplyr airbnb_listings %>%

top_n(3, number_of_rooms) Left Join

# Returns rows in left table and
missing values for any columns from the
Basic column operations with dplyr right table where joining field did not find a match

host_listings %>%

# Select one or more columns with select()

airbnb_listings %>%

> Aggregating data with dplyr left_join(airbnb_listings, by="listing_id")

select(listing_id, city) Right join

# Count groups within a column (e.g., count number of cities in airbnb_listings)

# Select columns based on start characters

airbnb_listings %>%
# Returns rows in right table and
missing values for any columns from the
count(city) left table where joining field did not find a match

airbnb_listings %>%

host_listings %>%

select(starts_with("c"))
# Count groups within a column and return sorted
right_join(airbnb_listings, by="listing_id")
# Select columns based on end characters
airbnb_listings %>%

airbnb_listings %>%
count(country, sort=TRUE)
Full Join
select(ends_with("s"))
# Return the total sum of values for a column (e.g., total number of rooms)
# Returns all records from both table, irrespective of whether there is a
# Select all but one column (e.g., listing_id)
airbnb_listings %>%
match on the joining field

airbnb_listings %>%
summarise(total_rooms=sum(number_of_rooms)) host_listings %>%

select(-listing_id) full_join(airbnb_listings, by="listing_id")

# Return the average of values for a column (e.g, average number of rooms in a given listing)

# Select all columns within a range

airbnb_listings %>%

summarise(avg_room=mean(number_of_rooms))
Anti Join
airbnb_listings %>%

select(country:year_listed) # Returns records in the first table and excludes matching values from the
# Return a custom summary statistic (e.g., average amount of time a listing stays on)
second table

# Reorder columns using relocate()

airbnb_listings %>%
airbnb_listings %>%

airbnb_listings %>%
summarise(average_listing_duration= 2022 - mean(year_listed)) anti_join(host_listings, by="listing_id")
relocate(city, country)
# Group by a variable and return counts of each group (e.g., number of listings by country)

# Rename a column using rename()

airbnb_listings %>%

airbnb_listings %>%
group_by(country) %>%

rename(year=year_listed) summarise(n=n())

# Select columns matching a regular expression

# Group by a variable and return the average value per group (e.g., average number of rooms
in listings per city)

airbnb_listings %>%

select(matches("(.n.)|(n.)")) airbnb_listings %>%

Learn Data Skills Online at www.DataCamp.com
group_by(city) %>%

summarise(avg_rooms=mean(number_of_rooms))