Grammar of data wrangling

wk3-d01-grammar-wrangle

Dr. D

Chico State
DATA 385 - Fall 2026

September 8, 2026

Setup

🎥 Grammar of wrangling - code along in wk3-d01

library(tidyverse)

A grammar of data wrangling…

… based on the concepts of functions as verbs that manipulate data frames

dplyr is part of the tidyverse

• select: pick columns by name
• arrange: reorder rows
• slice: pick rows using index(es)
• filter: pick rows matching criteria
• distinct: filter for unique rows
• mutate: add new variables
• summarise: reduce variables to values
• group_by: for grouped operations
• … (many more)

Rules of dplyr functions

• First argument is always a data frame
• Subsequent arguments say what to do with that data frame
• Always return a data frame
• Don’t modify in place

Reassure students: “don’t modify in place” means nothing is lost until you explicitly save over it. Feel free to be adventurous and try things — you can always recover, since you haven’t changed your original data frame unless you explicitly overwrite it.

Data: Hotel bookings

• Data from two hotels: one resort and one city hotel
• Observations: Each row represents a hotel booking
• Goal for original data collection: development of prediction models to classify a hotel booking’s likelihood to be cancelled (Antonio et al., 2019)

hotels <- read_csv("data/hotels.csv")

Source: TidyTuesday

First look: Variables

names(hotels)

 [1] "hotel"                          "is_canceled"                   
 [3] "lead_time"                      "arrival_date_year"             
 [5] "arrival_date_month"             "arrival_date_week_number"      
 [7] "arrival_date_day_of_month"      "stays_in_weekend_nights"       
 [9] "stays_in_week_nights"           "adults"                        
[11] "children"                       "babies"                        
[13] "meal"                           "country"                       
[15] "market_segment"                 "distribution_channel"          
[17] "is_repeated_guest"              "previous_cancellations"        
[19] "previous_bookings_not_canceled" "reserved_room_type"            
[21] "assigned_room_type"             "booking_changes"               
[23] "deposit_type"                   "agent"                         
[25] "company"                        "days_in_waiting_list"          
[27] "customer_type"                  "adr"                           
[29] "required_car_parking_spaces"    "total_of_special_requests"     
[31] "reservation_status"             "reservation_status_date"       

Second look: Overview

glimpse(hotels)

Rows: 119,390
Columns: 32
$ hotel                          <chr> "Resort Hotel", "Resort Hotel…
$ is_canceled                    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 1,…
$ lead_time                      <dbl> 342, 737, 7, 13, 14, 14, 0, 9…
$ arrival_date_year              <dbl> 2015, 2015, 2015, 2015, 2015,…
$ arrival_date_month             <chr> "July", "July", "July", "July…
$ arrival_date_week_number       <dbl> 27, 27, 27, 27, 27, 27, 27, 2…
$ arrival_date_day_of_month      <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
$ stays_in_weekend_nights        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ stays_in_week_nights           <dbl> 0, 0, 1, 1, 2, 2, 2, 2, 3, 3,…
$ adults                         <dbl> 2, 2, 1, 1, 2, 2, 2, 2, 2, 2,…
$ children                       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ babies                         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ meal                           <chr> "BB", "BB", "BB", "BB", "BB",…
$ country                        <chr> "PRT", "PRT", "GBR", "GBR", "…
$ market_segment                 <chr> "Direct", "Direct", "Direct",…
$ distribution_channel           <chr> "Direct", "Direct", "Direct",…
$ is_repeated_guest              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ previous_cancellations         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ previous_bookings_not_canceled <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ reserved_room_type             <chr> "C", "C", "A", "A", "A", "A",…
$ assigned_room_type             <chr> "C", "C", "C", "A", "A", "A",…
$ booking_changes                <dbl> 3, 4, 0, 0, 0, 0, 0, 0, 0, 0,…
$ deposit_type                   <chr> "No Deposit", "No Deposit", "…
$ agent                          <chr> "NULL", "NULL", "NULL", "304"…
$ company                        <chr> "NULL", "NULL", "NULL", "NULL…
$ days_in_waiting_list           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ customer_type                  <chr> "Transient", "Transient", "Tr…
$ adr                            <dbl> 0.00, 0.00, 75.00, 75.00, 98.…
$ required_car_parking_spaces    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ total_of_special_requests      <dbl> 0, 0, 0, 0, 1, 1, 0, 1, 1, 0,…
$ reservation_status             <chr> "Check-Out", "Check-Out", "Ch…
$ reservation_status_date        <date> 2015-07-01, 2015-07-01, 2015…

Select a single column

View only lead_time (number of days between booking and arrival date):

select(
  hotels,
  lead_time
)

# A tibble: 119,390 × 1
   lead_time
       <dbl>
 1       342
 2       737
 3         7
 4        13
 5        14
 6        14
 7         0
 8         9
 9        85
10        75
# ℹ 119,380 more rows

• Start with the function (a verb): select()
• First argument: data frame we’re working with, hotels
• Second argument: variable we want to select, lead_time
• Result: a data frame with 119390 rows and 1 column

Note

dplyr functions always expect a data frame and always yield a data frame.

Have students look at the actual lead_time values: 342 days and 737 days for the first two rows. Ask them to think back to pre-pandemic travel — how far ahead did they used to book hotels? 342 days is almost a year out, 737 is almost two years. Good quick engagement hook before dissecting the function call.

You try it: one column

Your turn

Select a single column of your choice — not lead_time. Before you run it, predict: how many rows and how many columns will the result have?

Select multiple columns

View only the hotel type and lead_time:

select(hotels, hotel, lead_time)

# A tibble: 119,390 × 2
   hotel        lead_time
   <chr>            <dbl>
 1 Resort Hotel       342
 2 Resort Hotel       737
 3 Resort Hotel         7
 4 Resort Hotel        13
 5 Resort Hotel        14
 6 Resort Hotel        14
 7 Resort Hotel         0
 8 Resort Hotel         9
 9 Resort Hotel        85
10 Resort Hotel        75
# ℹ 119,380 more rows

You try it: multiple columns

Your turn

Now select two or more columns of your choice. Same question — how many rows and how many columns do you expect this time?

Data wrangling, step-by-step

What if I wanted to sort the results by the lead time?

Select:

hotels %>%
  select(hotel, lead_time)

# A tibble: 119,390 × 2
   hotel        lead_time
   <chr>            <dbl>
 1 Resort Hotel       342
 2 Resort Hotel       737
 3 Resort Hotel         7
 4 Resort Hotel        13
 5 Resort Hotel        14
 6 Resort Hotel        14
 7 Resort Hotel         0
 8 Resort Hotel         9
 9 Resort Hotel        85
10 Resort Hotel        75
# ℹ 119,380 more rows

Select, then arrange:

hotels %>%
  select(hotel, lead_time) %>%
  arrange(desc(lead_time))

# A tibble: 119,390 × 2
   hotel        lead_time
   <chr>            <dbl>
 1 Resort Hotel       737
 2 Resort Hotel       709
 3 City Hotel         629
 4 City Hotel         629
 5 City Hotel         629
 6 City Hotel         629
 7 City Hotel         629
 8 City Hotel         629
 9 City Hotel         629
10 City Hotel         629
# ℹ 119,380 more rows

Note the introduction of a new operator called a “pipe”: %>%.

When narrating this, say “and then… and then…” out loud between the select and arrange steps — that verbal habit is literally the motivation for the pipe operator we’re about to introduce.

Pipes

What is a pipe?

In programming, a pipe is a technique for passing information from one process to another.

hotels %>%
  select(hotel, lead_time) %>%
  arrange(desc(lead_time))

# A tibble: 119,390 × 2
   hotel        lead_time
   <chr>            <dbl>
 1 Resort Hotel       737
 2 Resort Hotel       709
 3 City Hotel         629
 4 City Hotel         629
 5 City Hotel         629
 6 City Hotel         629
 7 City Hotel         629
 8 City Hotel         629
 9 City Hotel         629
10 City Hotel         629
# ℹ 119,380 more rows

• Start with the data frame hotels, and pass it to the select() function,
• then we select the variables hotel and lead_time,
• and then we arrange the data frame by lead_time in descending order.

Aside

The pipe operator %>% is implemented in the package magrittr, though we don’t need to load this package explicitly since tidyverse does this for us.

Recently (<2 yrs ago) Base R implemented it’s own pipe: |> that functions the exact same way as %>%. That allows you to use this feature without having to load a separate package.

And yes.. i’m sorry but this | vertical bar is also called a “pipe” in certain situations.

How does a pipe work?

You can think about the following sequence of actions: find keys, unlock car, start car, drive to work, park.

Expressed as a set of nested functions in R pseudocode this would look like:

park(drive(start_car(find("keys")), to = "work"))

Writing it out using pipes gives it a more natural (and easier to read) structure:

find("keys") %>%
  start_car() %>%
  drive(to = "work") %>%
  park()

A note on piping and layering

• %>% (or |>) used mainly in dplyr pipelines: we pipe the output of the previous line of code as the first input of the next line of code
• + used in ggplot2 plots is used for “layering”: we build the plot in layers, separated by +

dplyr uses %>%, not +

❌

hotels +
  select(hotel, lead_time)

Error:
! object 'hotel' not found

✅

hotels %>%
  select(hotel, lead_time)

# A tibble: 119,390 × 2
   hotel        lead_time
   <chr>            <dbl>
 1 Resort Hotel       342
 2 Resort Hotel       737
 3 Resort Hotel         7
 4 Resort Hotel        13
 5 Resort Hotel        14
 6 Resort Hotel        14
 7 Resort Hotel         0
 8 Resort Hotel         9
 9 Resort Hotel        85
10 Resort Hotel        75
# ℹ 119,380 more rows

The actual error here (“object ‘hotel’ not found”) is confusing on its face — it doesn’t obviously point to the real problem. Explain why: hotels never got passed into select() because + doesn’t pipe like %>% does, so select() is trying to evaluate hotel as a standalone object and can’t find it.

ggplot2 uses +, not %>%

❌

ggplot(hotels, 
       aes(x = hotel, 
           fill = deposit_type)) %>%
  geom_bar()

Error in `geom_bar()`:
! `mapping` must be created by `aes()`.
✖ You've supplied a <ggplot2::ggplot> object.
ℹ Did you use `%>%` or `|>` instead of `+`?

✅

ggplot(hotels, 
       aes(x = hotel, 
           fill = deposit_type)) +
  geom_bar()

Nice contrast with the dplyr case: ggplot2 is actually pretty good about this mistake — the error message will hint that you may have meant + instead of %>%. Worth pointing out that not all cryptic errors are equally unhelpful.

Code styling

Many of the styling principles are consistent across %>% and +:

• always a space before
• always a line break after (for pipelines with more than 2 lines)

❌

ggplot(hotels,aes(x=hotel,y=deposit_type))+geom_bar()

✅

ggplot(hotels, aes(x = hotel, y = deposit_type)) +
  geom_bar()

You try it: pipe it together

Your turn

Using the same columns you picked in “You try it: multiple columns,” write one pipe: hotels |> select(...) |> arrange(...).

Acknowledgements & Credits

Note

This page adapts material from Data Science in a Box (Unit 2, Deck 6: “Grammar of data wrangling”) by Mine Çetinkaya-Rundel, licensed under CC BY-SA 4.0. Source: tidyverse/datascience-box. Modified: converted from xaringan to Quarto revealjs. You try it’s added.