Day 1 - Girls in Data Science

Photos/Images by #WOCinTech/#WOCinTech Chat (CC-BY)

Team Introduction

• Lead Instructor: Katie Burak (Assistant Professor of Teaching, Department of Statistics, UBC)

• Data Science Mentors: Mona Zhu (PhD, MDS), Riya Eliza Shaju (MDS)

Thank you to our sponsors!

Camp Motivation

The main motivation for the development of this camp is to promote and uplift women and gender minorities in data science. The leaky pipeline is a phrase describing the phenomenon that even though women are interested in careers in STEM fields (including data science), they are disproportionately lost along the way as the level of their education and careers progresses. My hope is that camps like this one help to encourage and empower women to pursue careers in STEM fields, while exposing them to new and exciting topics in data science.

Ice Breaker Activity

• Write down an interesting fact about yourself (e.g., favourite hobby, a place you have travelled) and come and place your sticky on the white board (don’t write your name on the sticky note).

• Once everyone has placed a sticky note on the white board, come and take one that isn’t yours. It is your job to find out the name of the person whose random fact you found and return their sticky note to them!

• After everyone has their original note, please share your fun facts with everyone at your table.

Day 1 Learning Objectives

• Edit and execute code and markdown in Jupyter notebooks

• Understand the essentials of the R programming language

• Apply the basics of data manipulation and data visualization in R

We’ve got a lot to do - let’s get started!

1. Getting started with Jupyter & R

This webpage is called a Jupyter notebook. A notebook is a place to write computer code for analysis, view the results of the analysis, as well as to narrate the analysis with rich formatted text.

1.1. Text Cells

In a notebook, each rectangle containing text or code is called a cell.

Text cells (like this one) can be edited by double-clicking on them. They’re written in a simple format called Markdown to add formatting and section headings. For more practice using Markdown, try this online tutorial!

After you edit a text cell, click the “run cell” button at the top that looks like ▶ to confirm any changes. (Try not to delete the instructions of the lab.)

1.2. Code Cells

Other cells contain code in the R language. Running a code cell will execute all of the code it contains.

To run the code in a cell, first click on that cell to activate it. It will be highlighted with a blue rectangle to the left of it when activated. Next, either press Run ▶ or hold down the shift key and press return or enter. print(“Hello, World!”) Try running the next cell:

print("Hello, World!")

[1] "Hello, World!"

The above code cell contains a single line of code, but cells can also contain multiple lines of code. When you run a cell, the lines of code are executed in the order in which they appear. Every print expression prints a line. Run the next cell and notice the order of the output.

print("First this line is printed,")
print("and then this one.")

[1] "First this line is printed,"

[1] "and then this one."

1.3. Writing Jupyter Notebooks

You can use Jupyter notebooks for your own projects or documents. When you make your own notebook, you’ll need to create your own cells for text and code.

To add a cell, click the + button in the menu bar of this tab. The newly created cell will start out as a code cell. You can change it to a text cell by clicking inside it so it’s highlighted, clicking the drop-down box next to the restart and run all button (⏩) in the menu bar of this tab, and changing it from “Code” to “Markdown”.

Exercise

Add a code cell below this one. Write code in it that prints out:

My name is ____!

Run your cell to verify that it works.

Next, add a text/Markdown cell and write the text “My name is ____!” in it.

1.4. Comments

Below you see lines like this in code cells:

# Test cell; please do not change!

That is called a comment. It doesn’t make anything happen in R; R ignores anything on a line after a #. Instead, it’s there to communicate something about the code to you, the human reader. Comments are extremely useful and can help increase how readable our code is.

The below code cell contains comments (one at the start of a line, and one after some other code). Run the cell. You will see that everything after a comment symbol # is ignored by R.

# you can use comments to document your code, or make R ignore some code without deleting it entirely
# print("this is a commented line that R will ignore. You won't see this text in the output!")

print("hello!") # you can also put comments at the end of a line of code

[1] "hello!"

1.5. Errors

R is a language, and like natural human languages, it has rules. It differs from natural language in two important ways:

1. The rules are simple. You can learn most of them in a few weeks and gain reasonable proficiency with the language in a semester.

2. The rules are rigid. If you’re proficient in a natural language, you can understand a non-proficient speaker, glossing over small mistakes. A computer running R code is not smart enough to do that.

Whenever you write code, you’ll make mistakes (everyone who writes code does, even your course instructor!). When you run a code cell that has errors, R will sometimes produce error messages to tell you what you did wrong.

Errors are totally okay; even experienced programmers make many errors. It’s a natural part of the coding process. When you make an error, you just have to find the source of the problem, fix it, and move on.

We have made an error in the next cell. Remove the # symbol below (i.e., uncomment the line), and then run the cell to see what happens.

# print("This line is missing something."

There’s a lot of terminology in programming languages, but you don’t need to know it all in order to program effectively. Even though the error message can seem cryptic, if you read it carefully you’ll often find hints as to what went wrong. For example, above, you’ll see the message unexpected end of input (among a lot of other technical jargon). In other words, R reached the end of the line of code, and wasn’t expecting to reach the end – it thinks there is still something missing!

Of course, even if you do your best to interpret the error message, sometimes you may get stuck figuring out what went wrong and how to fix it. In that case, ask someone.

Try to fix the code above so that you can run the cell and see the intended message instead of an error.

1.6 Saving your work

Its important to save your work often so you don’t lose your progress! At the top of the screen, go to the File menu then Save Notebook. Also, there are keyboard shorcuts for saving your work too: control + s on Windows, or command + s on Mac. Once you’ve saved your work, you will see a message at the bottom of the screen that says Saving completed.

1.7 Numbers

Quantitative information arises everywhere in data science. In addition to representing commands to print out lines, our R code can represent numbers and methods of combining numbers. The expression 3.2500 evaluates to the number 3.25. (Run the cell and see.)

3.2500

3.25

Notice that we didn’t have to write print(). When you run a notebook cell, Jupyter helpfully prints out that value for you.

2
3
4

2

3

4

Above, you should see that the three numbers (2, 3, and 4) are printed out. In R, simply inputting numbers and running the cell will generate all the numbers that you listed.

1.8 Arithmetic

The line in the next cell performs some mathemtical operations. Run them!

2.0 - 1.5

0.5

2 * 2

4

6/2

3

Many basic arithmetic operations are built in to R. This webpage describes many arithmetic operators that might be useful in this camp.

1.9 Names

In natural language, we have terminology that lets us quickly reference very complicated concepts. We don’t say, “That’s a large mammal with brown fur and sharp teeth!” Instead, we just say, “Bear!”

Similarly, an effective strategy for writing code is to define names for data as we compute it, like a lawyer would define terms for complex ideas at the start of a legal document to simplify the rest of the writing.

In R, we do this with objects. An object has a name on the left side of an <- sign and an expression to be evaluated on the right.

answer <- 3*2 + 4

When you run that cell, R first evaluates the first line. It computes the value of the expression 3 * 2 + 4, which is the number 10. Then it gives that value the name answer. At that point, the code in the cell is done running.

After you run that cell, the value 10 is bound to the name answer:

answer

10

Note: You can also use an = sign for assignment

When naming objects in R there are some rules:

1. Names in R can have letters (upper- and lower-case letters are both okay and count as different letters e.g. “Answer” and “answer” will be treated as different objects), underscores, dots, and numbers.

2. The first character can’t be a number (otherwise a name might look like a number).

3. Names can’t contain spaces, since spaces are used to separate pieces of code from each other.

Other than those rules, what you name something doesn’t matter to R. For example, the next cell does the same thing as the above cell, except everything has a different name:

a <- 840
b <- 2 * a
c <- 12
d <- c * b
d

20160

Another common pattern is that a series of lines in a single cell will build up a complex computation in stages, naming the intermediate results.

However, names are very important for making your code readable to yourself and others. The cell above is shorter, but it’s totally useless without an explanation of what it does.

There is also cultural style associated with different programming languages. In the modern R style, object names should use only lowercase letters, numbers, and _. Underscores (_) are typically used to separate words within a name (e.g., answer_one).

Exercise

Question: How old will you be in 2050?

• Assign an object called age which is your age now.

• Compute how many years will pass until 2050 by subtracting the current year from 2050 (call this object years).

• Finally, name an object future_age which adds together age and years to get your age in 2050.

1.10 Functions

The most common way to combine or manipulate values in R is by calling functions. R comes with many built-in functions that perform common operations.

We used a function print() at the beginning of this notebook when we printed text from a code cell. Here we’ll demonstrate using another function toupper() that converts text to uppercase:

greeting <- toupper("Why, hello there!")
greeting

'WHY, HELLO THERE!'

Try using the function tolower now!

Some functions take multiple arguments, separated by commas. For example, the built-in max function returns the maximum argument passed to it.

biggest <- max(4,7,23,1,5)
biggest

23

Try to use the min function now to find the smallest number!

1.11 Packages

R has many built-in functions, but we can also use functions that are stored within packages created by other R users. We are going to use a package, called tidyverse, to load, modify and plot data. This package has already been installed for you. Later in the course you will learn how to install packages so you are free to bring in other tools as you need them for your data analysis.

To use the functions from a package you first need to load it using the library function. This needs to be done once per notebook (and a good rule of thumb is to do this at the very top of your notebook so it is easy to see what packages your R code depends on).

library(tidyverse)

── Attaching core tidyverse packages ─────────────────────────────────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.4     ✔ readr     2.1.5
✔ forcats   1.0.0     ✔ stringr   1.5.1
✔ ggplot2   3.5.1     ✔ tibble    3.2.1
✔ lubridate 1.9.3     ✔ tidyr     1.3.1
✔ purrr     1.0.2     

── Conflicts ───────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

Note: it is normal and expected that a message is printed out after loading the tidyverse and some packages. Generally, this message let’s you know if functions from the different packages were loaded share the same name (which is confusing to R), and if so, which one you can access using just it’s name (and which one you need to refer the package name and the function name to refer to it, this is called masking). Additionally, the tidyverse is a special R package - it is a meta-package that bundles together several related and commonly used packages. Because of this it lists the packages it does the job of loading.

1.12 Looking for Help

No one, even experienced, professional programmers remember what every function does, nor do they remember every possible function argument/option. So both experienced and new programmers (like you!) need to look things up, A LOT!

One of the most efficient places to look for help on how a function works is the R help files. Let’s say we wanted to pull up the help file for the toupper() function. We can do this by typing a question mark in front of the function we want to know more about. Remove the hashtag and run the cell below to find out more about toupper().

# ?toupper

At the very top of the file, you will see the function itself and the package it is in (in this case, it is base). Next is a description of what the function does. You’ll find that the most helpful sections on this page are “Usage”, “Arguments” and “Examples”.

• Usage gives you an idea of how you would use the function when coding–what the syntax would be and how the function itself is structured.

• Arguments tells you the different parts that can be added to the function to make it more simple or more complicated. Often the “Usage” and “Arguments” sections don’t provide you with step by step instructions, because there are so many different ways that a person can incorporate a function into their code. Instead, they provide users with a general understanding as to what the function could do and parts that could be added. At the end of the day, the user must interpret the help file and figure out how best to use the functions and which parts are most important to include for their particular task.

• The Examples section is often the most useful part of the help file as it shows how a function could be used with real data. It provides a skeleton code that the users can work off of.

Beyond the R help files there are many resources that you can use to find help. Stack overflow, an online forum, is a great place to go and ask questions such as how to perform a complicated task in R or why a specific error message is popping up. Oftentimes, a previous user will have already asked your question of interest and received helpful advice from fellow R users.

2. Introduction to Data Science

What is data science exactly?

Data science is the use of reproducible and auditable processes to obtain value (i.e., insight) from data.

Every good data analysis begins with a question—like the above—that you aim to answer using data. As it turns out, there are actually a number of different types of question regarding data:

• descriptive

• exploratory

• predictive

• inferential

• causal

• mechanistic.

Note: In this camp, we will focus on the first 3 types of questions.

• Descriptive: A question which asks about summarized characteristics of a data set without interpretation (i.e., report a fact, describe characteristics)

• Exploratory: A question asks if there are patterns, trends, or relationships within a single data set. Often used to propose hypotheses for future study. (discovery of ideas and thoughts)

• inferential: determine if association observed in your exploratory analysis hold in a different sample that is representative of a population

• predictive: what predicts what grade someone will achieve in a certain class

• causal: whether changing one factor will change another factor

• mechanistic: how e.g. how a medication leads to a reduction in the number of illnesses

Why are we using programming languages to do data analysis?

VS.

• There are many advantages to using R (or another language, like Python or Julia):

  • statistical analysis functions that go beyond Excel

  • free and open-source

  • transparent & reproducible code

  • can handle large amounts of data and complex analyses

• Using a programming language is like baking with a recipe:

  • Ingredients = data

    

  • Recipe = code

    

• Someone else can use your recipe (code) to bake the same cake (produce the same data analyses)

• Spreadsheets in Excel make it very difficult to understand where results came from

In the data science workflow (source: Grolemund & Wickham, R for Data Science)

2.1 Reading in Tabular Data

Loading/importing data

• 4 most common ways to do this in Data Science

  1. read in a text file with data in a spreadsheet format

  2. read from a database (e.g., SQLite, PostgreSQL)

  3. scrape data from the web

  4. use a web API (Application Programming Interface) to read data from a website

We will focus on option 1!

It is important to read in data carefully and check results after! This will help reduce bugs and speed up your analyses down the road… think of it as tying your shoes before you run; not exciting, but if done wrong it will trip you up later!

Different ways to locate a file / dataset

Local (on your computer)

• An absolute path locates a file with respect to the “root” folder on a computer

  • starts with /

    e.g. /home/instructor/documents/timesheet.xlsx

• A relative path locates a file relative to your working directory

  • doesn’t start with /

    e.g. documents/timesheet.xlsx
    (where working directory is /home/instructor/)

Remote (on the web)

via “URL” that starts with http:// or https://

http://traffic.libsyn.com/mbmbam/MyBrotherMyBrotherandMe367.mp3

Absolute vs relative paths: Which should you use?

• Generally to ensure your code can be run on a different computer, you should use relative paths

• e.g. Alice is working inside the folder /home/alice/project/. To specify where to load data from in her Jupyter notebook, she uses the absolute path /home/alice/project/data/happiness_report.csv. What issue will arise when she shares the notebook with her collaborator Keeran who tries to read in the data on their computer?

/home/keeran/project/data/happiness_report.csv

What relative data path could they use to collaborate more effectively?

data/happiness_report.csv

• Even though stored their files in the same place on their computers (in their home folders), the absolute paths are different due to their different usernames.

• If Alice has code that loads the happiness_report.csv data using an absolute path, the code won’t work on Keeran’s computer. But the relative path from inside the project folder (data/happiness_report.csv) is the same on both computers; any code that uses relative paths will work on both.

• Abosolute paths are like GPS coordinates, they take you to one specific location regardless of where you are starting from. Relative paths are like directions, they are based off your starting point (e.g. go to blocks north and then one west).

A data set is a structured collection of numbers and characters. Aside from that, there are really no strict rules; data sets can come in many different forms! Perhaps the most common form of data set that you will find in the wild, however, is tabular data. Think spreadsheets in Microsoft Excel: tabular data are rectangular-shaped and spreadsheet-like.

When we load tabular data into R, it is represented as a data frame object. We refer to the rows as observations and columns as variables.

The main kind of data file that we will learn how to load into R as a data frame is the comma-separated values format (.csv for short). These files have names ending in .csv, and can be opened and saved using common spreadsheet programs like Microsoft Excel and Google Sheets.

To load data into R so that we can do things with it (e.g., perform analyses or create data visualizations), we will need to use a function. A function is a special word in R that takes instructions (we call these arguments) and does something. The function we will use to load a .csv file into R is called read_csv (made accessible by loading the tidyverse R package). In its most basic use-case, read_csv expects that the data file:

• has column names (or headers)

• uses a comma (,) to separate the columns

• does not have row names

Please note that data comes in many forms and there are a wide variety of functions and approaches to loading in your data, but in this camp we will be focusing on reading in tabular data using read_csv.

We will now look at an Instagram data set focusing on 200 of the most popular instagram accounts. Let’s try reading it in with read_csv using a relative path!

insta <- read_csv('data/insta.csv')

insta <- insta |> select(-'Avg. Likes')

head(insta) # head function prints only the first 6 rows

Rows: 200 Columns: 8

── Column specification ───────────────────────────────────────────────────────────────────────────────────
Delimiter: ","
chr (7): name, channel_Info, Category, Posts, Followers, Avg. Likes, Eng Rate
dbl (1): rank

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

A tibble: 6 × 7

rank	name	channel_Info	Category	Posts	Followers	Eng Rate
<dbl>	<chr>	<chr>	<chr>	<chr>	<chr>	<chr>
1	instagram	brand	photography	7.3K	580.1M	0.1%
2	cristiano	male	Health, Sports & Fitness	3.4K	519.9M	1.4%
3	leomessi	male	Health, Sports & Fitness	1K	403.7M	1.7%
4	kyliejenner	female	entertainment	7K	375.9M	1.7%
5	selenagomez	female	entertainment	1.8K	365.3M	1.1%
6	therock	male	entertainment	7K	354.3M	0.3%

Run the following code chunk before continuing to rename/reformat some of the variables:

insta <- suppressWarnings(insta |>
              mutate(Followers = as.numeric(str_replace(Followers, "M", ""))*1e6) |> 
              rename(Channel = channel_Info, eng_rate = 'Eng Rate') |> 
             mutate(
                    Posts = case_when(
                      str_detect(Posts , "K") ~ as.numeric(str_replace(Posts , "K", "")) * 1000,
                      str_detect(Posts, "M") ~ as.numeric(str_replace(Posts, "M", "")) * 1000000,
                      TRUE ~ as.numeric(Posts)  # If no suffix, just convert to numeric
                    )
                  ) |>               
             mutate(eng_rate = as.numeric(str_replace(eng_rate, "%", ""))) |> 
             mutate(Category = if_else(is.na(Category), "Not Available", Category))|> 
             mutate(Channel = if_else(is.na(Channel), "Not Available", Channel)))

insta

A tibble: 200 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
1	instagram	brand	photography	7300	580100000	0.1
2	cristiano	male	Health, Sports & Fitness	3400	519900000	1.4
3	leomessi	male	Health, Sports & Fitness	1000	403700000	1.7
4	kyliejenner	female	entertainment	7000	375900000	1.7
5	selenagomez	female	entertainment	1800	365300000	1.1
6	therock	male	entertainment	7000	354300000	0.3
7	arianagrande	female	entertainment	5000	345600000	1.4
8	kimkardashian	female	entertainment	5700	336300000	0.9
9	beyonce	female	entertainment	2100	287300000	1.0
10	khloekardashian	female	entertainment	4200	283900000	0.5
11	justinbieber	male	entertainment	7400	270200000	0.5
12	nike	brand	Health, Sports & Fitness	1000	257600000	0.1
13	taylorswift	male	entertainment	562	236200000	1.3
14	jlo	female	entertainment	220	228800000	0.5
15	virat.kohli	male	Not Available	1500	228000000	1.2
16	kendalljenner	male	Not Available	731	223400000	2.3
17	natgeo	brand	entertainment	26700	220600000	0.1
18	nickiminaj	female	entertainment	6400	206800000	0.8
19	kourtneykardash	female	entertainment	4400	206200000	0.8
20	kendalljenner	male	Not Available	824	204400000	2.5
21	neymarjr	male	Health, Sports & Fitness	5400	196200000	1.5
22	natgeo	male	Not Available	26000	196100000	0.1
23	mileycyrus	female	entertainment	1200	189900000	0.5
24	katyperry	female	entertainment	2100	179800000	0.3
25	zendaya	female	entertainment	3500	161800000	4.1
26	kevinhart4real	male	entertainment	8400	157600000	0.2
27	iamcardib	male	entertainment	1600	145600000	1.8
28	ddlovato	female	entertainment	68	143500000	0.3
29	badgalriri	female	entertainment	4900	139200000	2.7
30	kingjames	male	Health, Sports & Fitness	2400	139200000	0.9
⋮	⋮	⋮	⋮	⋮	⋮	⋮
171	parineetichopra	female	entertainment	1400	39200000	0.9
172	anushkasen0408	female	entertainment	5500	39000000	1.5
173	chrissyteigen	female	entertainment	5100	38600000	0.7
174	hm	brand	fashion	7500	38500000	0.1
175	wesleysafadao	male	entertainment	9000	38300000	0.3
176	marvelstudios	community	entertainment	2800	38200000	1.3
177	houseofhighlights	male	Health, Sports & Fitness	26900	37900000	0.8
178	tyga	Not Available	entertainment	27	37700000	1.7
179	eminem	male	entertainment	721	37600000	2.4
180	sachintendulkar	male	Health, Sports & Fitness	1100	37500000	1.7
181	danialves	male	Health, Sports & Fitness	3400	37400000	0.8
182	gisel_la	female	entertainment	10200	37400000	0.2
183	blakelively	male	entertainment	121	37000000	5.6
184	chelseafc	community	Health, Sports & Fitness	17300	36600000	0.6
185	shahidkapoor	male	entertainment	1200	36600000	2.4
186	kimberly.loaiza	female	entertainment	578	36600000	4.4
187	toni.kr8s	male	Health, Sports & Fitness	1000	36600000	1.2
188	antogriezmann	male	Health, Sports & Fitness	872	36400000	1.8
189	mercedesbenz	brand	technology	18500	36400000	0.3
190	nattinatasha	female	entertainment	54	36100000	1.3
191	tigerjackieshroff	male	entertainment	2200	36100000	1.9
192	lunamaya	female	entertainment	4200	36100000	0.2
193	mancity	male	Health, Sports & Fitness	19600	36000000	0.4
194	disney	brand	entertainment	7700	35800000	0.3
195	barackobama	male	News & Politics	743	35500000	1.7
196	fcbayern	male	Health, Sports & Fitness	16800	35400000	0.6
197	colesprouse	male	entertainment	1100	35300000	3.5
198	shaymitchell	male	entertainment	6300	35100000	1.2
199	ivetesangalo	female	entertainment	7800	35000000	0.4
200	paollaoliveirareal	female	entertainment	4800	34900000	0.7

It looks like this data set has 200 rows (observations) representing the top 200 instgram accounts and the following 8 columns (variables):

• rank (1-200)

• name (Instagram handle)

• channel (brief description of the account)

• category

• posts

• followers

• Eng Rate (calculates the account’s engagement rate by dividing the total number of likes and comments received by the total number of followers, expressed as a percentage).

Without looking at the entire data set, we can find out attributes of our data such as the number of rows, columns, or overall dimension.

# Number of rows
nrow(insta)

# Number of columns
ncol(insta)

# Dimension 
dim(insta)

200

7

1. 200
2. 7

2.2 Data Wrangling!

The cartoon illustrations are by Allison Horst

• In the real world, when you get data, it’s usually very messy

  • inconsistent format (commas, tabs, semicolons, missing data, extra empty lines)

  • split into multiple files (e.g. yearly recorded data over many years)

  • corrupted files, custom formats

• when you read it successfully into R, it will often still be very messy

• you need to make your data “tidy”

What is Tidy Data?

Illustrations from the Openscapes blog Tidy Data for reproducibility, efficiency, and collaboration by Julia Lowndes and Allison Horst”

True or False: The Instagram data set is tidy

True!

• each row corresponds to a single observation,

• each column corresponds to a single variable, and

• each cell (row, column pair) correspond to a single value

Tools for tidying and wrangling data

• tidyverse package functions from:

  • dplyr package (select, filter, mutate, group_by, summarize)

2.1.1 Mutate

The mutate function transforms old columns to add new columns.

e.g. convert engagement rate to a decimal

head(mutate(insta, eng_rate = eng_rate / 100)) # head function returns only the first 6 rows

A tibble: 6 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
1	instagram	brand	photography	7300	580100000	0.001
2	cristiano	male	Health, Sports & Fitness	3400	519900000	0.014
3	leomessi	male	Health, Sports & Fitness	1000	403700000	0.017
4	kyliejenner	female	entertainment	7000	375900000	0.017
5	selenagomez	female	entertainment	1800	365300000	0.011
6	therock	male	entertainment	7000	354300000	0.003

Note: The above creates a new dataframe, it does not save it to the original insta df. We would need to assign it to a new variable if we want to save it.

2.1.2 Select

The select function is used to select a subset of columns (variables) from a dataframe.

head(select(insta, name)) # select `name` column 

A tibble: 6 × 1

name
<chr>
instagram
cristiano
leomessi
kyliejenner
selenagomez
therock

head(select(insta, name, Followers)) # select `name` and `Followers` column 

A tibble: 6 × 2

name	Followers
<chr>	<dbl>
instagram	580100000
cristiano	519900000
leomessi	403700000
kyliejenner	375900000
selenagomez	365300000
therock	354300000

2.1.3 Filter

The filter function is used to choose a subset of rows (observations) in a dataframe.

e.g. filter to only include instagram accounts with more than 150 million followers

# One condition
filter(insta, Followers > 150000000)

A tibble: 26 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
1	instagram	brand	photography	7300	580100000	0.1
2	cristiano	male	Health, Sports & Fitness	3400	519900000	1.4
3	leomessi	male	Health, Sports & Fitness	1000	403700000	1.7
4	kyliejenner	female	entertainment	7000	375900000	1.7
5	selenagomez	female	entertainment	1800	365300000	1.1
6	therock	male	entertainment	7000	354300000	0.3
7	arianagrande	female	entertainment	5000	345600000	1.4
8	kimkardashian	female	entertainment	5700	336300000	0.9
9	beyonce	female	entertainment	2100	287300000	1.0
10	khloekardashian	female	entertainment	4200	283900000	0.5
11	justinbieber	male	entertainment	7400	270200000	0.5
12	nike	brand	Health, Sports & Fitness	1000	257600000	0.1
13	taylorswift	male	entertainment	562	236200000	1.3
14	jlo	female	entertainment	220	228800000	0.5
15	virat.kohli	male	Not Available	1500	228000000	1.2
16	kendalljenner	male	Not Available	731	223400000	2.3
17	natgeo	brand	entertainment	26700	220600000	0.1
18	nickiminaj	female	entertainment	6400	206800000	0.8
19	kourtneykardash	female	entertainment	4400	206200000	0.8
20	kendalljenner	male	Not Available	824	204400000	2.5
21	neymarjr	male	Health, Sports & Fitness	5400	196200000	1.5
22	natgeo	male	Not Available	26000	196100000	0.1
23	mileycyrus	female	entertainment	1200	189900000	0.5
24	katyperry	female	entertainment	2100	179800000	0.3
25	zendaya	female	entertainment	3500	161800000	4.1
26	kevinhart4real	male	entertainment	8400	157600000	0.2

Note the usage of two equals signs ==. In coding, two equals signs evaluates equality between two values and returns either TRUE or FALSE. This is different than the use of one equals sign which is used to assign values.

# Two conditions
filter(insta, Followers > 150000000 & Category == "Health, Sports & Fitness")

A tibble: 4 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
2	cristiano	male	Health, Sports & Fitness	3400	519900000	1.4
3	leomessi	male	Health, Sports & Fitness	1000	403700000	1.7
12	nike	brand	Health, Sports & Fitness	1000	257600000	0.1
21	neymarjr	male	Health, Sports & Fitness	5400	196200000	1.5

### Example ###

a = "Pancakes" # assigns the value "Pancakes" to "a" 

a == "Pancakes" # returns TRUE

a == "Waffles" # returns FALSE

TRUE

FALSE

You can also use the | symbol in place of the word “or”. For example:

filter(insta, Category == "photography" | Category == "fashion")

A tibble: 12 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
1	instagram	brand	photography	7300	580100000	0.1
49	gigihadid	female	fashion	3300	76400000	3.4
51	victoriassecret	brand	fashion	3100	73800000	0.1
93	zara	brand	fashion	3900	55200000	0.1
113	louisvuitton	brand	fashion	6600	50100000	0.2
116	gucci	brand	fashion	9200	49800000	0.2
134	natgeotravel	brand	photography	17200	46000000	0.2
148	uarmyhope	Not Available	fashion	171	42200000	18.5
149	georginagio	female	fashion	778	42200000	5.7
153	voguemagazine	brand	fashion	11000	41800000	0.3
156	virginia	female	fashion	2000	41400000	4.0
174	hm	brand	fashion	7500	38500000	0.1

2.1.4 Slice

The slice function allows us to extract certain rows from our data set. For example, we may want to look at the 60th row of our data set (corresponding to the 60th most followed instagram account):

slice(insta, 60)

A tibble: 1 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
60	emmawatson	female	entertainment	417	68800000	1.1

You could also slice over a range of rows, say rows 60-70.

slice(insta, 60:70)

A tibble: 11 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
60	emmawatson	female	entertainment	417	68800000	1.1
61	ronaldinho	male	Health, Sports & Fitness	3000	68300000	1.0
62	tomholland2013	male	entertainment	1200	67700000	9.9
63	justintimberlake	male	entertainment	790	67500000	0.7
64	marvel	community	entertainment	7400	67000000	0.4
65	sooyaaa__	female	entertainment	900	66600000	8.0
66	raffinagita1717	male	entertainment	18300	66000000	0.4
67	camila_cabello	female	entertainment	3000	65800000	1.8
68	roses_are_rosie	female	entertainment	879	65200000	8.3
69	jacquelinef143	female	entertainment	2500	64300000	1.1
70	akshaykumar	male	entertainment	2000	63600000	2.0

2.1.4 Pipe Operator

When you need to type out a long sequence of operations on data, you could either:

(A) Save intermediate objects

(B) Compose function

(C) Use the pipe operator |>

Let’s look closer at each one of these:

(A) Save intermediate objects

insta_1 <- mutate(insta, eng_rate = eng_rate / 100)
insta_2 <- filter(insta_1, Followers > 150000000)
insta_3 <- select(insta_2, name, Category, Followers, eng_rate)
insta_3

Disadvantages:

• The reader may be tricked into thinking the named insta_1 and insta_2 objects are important for some reason, while they are just temporary intermediate computations.

• Further, the reader has to look through and find where insta_1 and insta_2 are used in each subsequent line.

• Creating variables that we don’t need, could lead to memory issues if they are big and take up a lot of space.

(B) Composing functions

insta_composed <- select(
    filter(
        mutate(
            insta, eng_rate = eng_rate / 100
        ),
        Followers > 150000000
    ),
   name, Category, Followers, eng_rate
)
insta_composed

Disadvantage:

• Difficult to read since you the innermost function is exectued first

• You have to start reading from the middle of the code block (mutate above)

(C) Piping

You can also pipe with the |> symbol: passes the output of a function to the 1st argument of another.

insta_piped <- insta |>
    mutate(insta, eng_rate = eng_rate / 100) |>
    filter(Followers > 150000000) |>
    select(name, Category, Followers, eng_rate)
insta_piped

note: if R sees a |> at the end of a line, it keeps reading the next line of code before evaluating!

Advantage:

• Pipes make code much more readable when you need to do a long sequence of operations on data.

• No intermediatery variables created

insta_piped <- insta |>
    mutate(insta, eng_rate = eng_rate / 100) |>
    filter(Followers > 150000000) |>
    select(name, Category, Followers, eng_rate)

head(insta_piped)

A tibble: 6 × 4

name	Category	Followers	eng_rate
<chr>	<chr>	<dbl>	<dbl>
instagram	photography	580100000	0.001
cristiano	Health, Sports & Fitness	519900000	0.014
leomessi	Health, Sports & Fitness	403700000	0.017
kyliejenner	entertainment	375900000	0.017
selenagomez	entertainment	365300000	0.011
therock	entertainment	354300000	0.003

2.1.5 Grouping and summarizing via group_by + summarize

• Grouping is when you split data into groups based on the value of a column.

• Sumarizing is when you combine data into fewer summary values.

For example computing the average particle pollution per city (source https://info201.github.io/dplyr.html):

• Another example, splitting the insta data into one group per species, and then summarizing the values within each group e.g. reporting the average Followers per account category.

• To do this we use group_by + summarize to iterate over Category, calculating average followers.

insta |> group_by(Category) |> summarize(avg_followers = mean(Followers))

• You might be interested in arranging this dataframe from smallest to largest number of followers.

• The arrange function allows us to order the rows of a data frame by the values of a particular column (in descending order, for example):

insta |> 
    group_by(Category) |>
    summarize(avg_followers = mean(Followers)) |> 
    arrange(by = desc(avg_followers))

A tibble: 12 × 2

Category	avg_followers
<chr>	<dbl>
photography	313050000
Not Available	140628571
Health, Sports & Fitness	88161538
entertainment	84184496
technology	61100000
Lifestyle	58100000
News & Politics	52000000
Finance	51900000
fashion	51140000
Beauty & Makeup	49133333
food	46900000
Craft/DIY	46200000

Or we could count the number of accounts in each category using n():

insta |> 
    group_by(Category) |>
    summarize(Count = n()) |> 
    arrange(by = desc(Count))

A tibble: 12 × 2

Category	Count
<chr>	<int>
entertainment	129
Health, Sports & Fitness	39
fashion	10
Not Available	7
Beauty & Makeup	3
News & Politics	3
food	2
photography	2
technology	2
Craft/DIY	1
Finance	1
Lifestyle	1

Exercise

Let’s practice what we’ve learned and try to answer the question “Which category of brand accounts have the most total Posts?” Using the pipe operator, return a dataframe based on the following criteria:

• Include the name of the account, the channel information, the category, and how many posts they have made

• Filter to include only accounts that are brands

• Grouping by category, compute the average total posts (call it avg_posts) and arrange the dataframe in descending order based on the average posts.

• Finally, make sure to assign this wrangled data to an object called insta_wrangled.

### YOUR CODE HERE

3. Data Visualization

Attribution: images that are not accompanied by code mostly come from
The Fundamentals of Data Visualization by Claus O. Wilke

Artwork by @allison_horst

Designing a visualization: ask a question, then answer it

The purpose of a visualization is to answer a question about a dataset of interest.

A good visualization answers the question clearly. A great visualization also hints at the question itself.

Visualizations alone help us answer two types of questions:

• descriptive: What are the top 10 most followed instagram accounts?

• exploratory: Is there a relationship between number of posts and followers?

• ~~inferential~~

• ~~predictive~~

• ~~causal~~

• ~~mechanistic~~

(we need more tools + visualizations to answer the others)

3.1 Creating visualizations in R

• It’s an iterative procedure. Try things, make mistakes, and refine!

• We will use ggplot2. There are three key aspects of plots in ggplot2:

  1. aesthetic mappings: map dataframe columns to visual properties

  2. geometric objects: encode how to display those visual properties

  3. scales: transform variables, set limits

• Add these one by one using +

Check out this R4DS reading for more details on data visualization with ggplot2.

Types of variables

A variable refers to a characteristic of interest and can be:

1. categorical: can be divided into groups (categories) e.g. marital status

2. quantitative: measured on a numeric scale (usually units are attached) e.g. height

Exercise

Looking at the Instgram data set, identify each variable as either categorical or quantitative.

3.1.1 Histograms

To visualize the distribution of a single quantitative variable

e.g. What does the distribution of the number of posts look like?

# Inspect the data
head(insta)

A tibble: 6 × 7

rank	name	Channel	Category	Posts	Followers	eng_rate
<dbl>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>
1	instagram	brand	photography	7300	580100000	0.1
2	cristiano	male	Health, Sports & Fitness	3400	519900000	1.4
3	leomessi	male	Health, Sports & Fitness	1000	403700000	1.7
4	kyliejenner	female	entertainment	7000	375900000	1.7
5	selenagomez	female	entertainment	1800	365300000	1.1
6	therock	male	entertainment	7000	354300000	0.3

ggplot(insta, aes(x = Posts)) +
    geom_histogram(bins=50) + 
    theme(text = element_text(size = 26)) + # increase text size
    labs(x='Number of Posts', y='Count', title='Histogram of Posts') # rename axes and add title

You can also creating multiple separate histograms (e.g., for each channel):

# Set the default size for plot
options(repr.plot.width = 8, repr.plot.height = 10)
ggplot(insta, aes(x = Posts, fill = Channel)) +
    geom_histogram(bins=50) +
    facet_grid(rows = vars(Channel)) +
    theme(text = element_text(size = 26)) 

3.1.2 Scatter Plots

To visualize the relationship between two quantitative variables

e.g. Is there a relationship between number of posts and number of followers?

# Set the default size for plot
options(repr.plot.width = 10, repr.plot.height = 8)

ggplot(insta, aes(x = Posts, y = Followers)) +
    geom_point() +
    theme(text = element_text(size = 26)) # increase text size

Let’s take a look at accounts that have over 10000 posts.

insta_posts_10000 <- insta |>
                        filter(Posts>10000)

# Which of these accounts has the most followers?
insta_posts_10000$name[which.max(insta_posts_10000$Followers)]

'natgeo'

options(repr.plot.width = 10, repr.plot.height = 8)

ggplot(insta_posts_10000, aes(x = Posts, y = Followers, color = Channel)) + # color by channel info
    geom_point(size=4) +
    theme(text = element_text(size = 26)) +
    geom_text(x = 45000, y = 221000000, label = "National Geographic", color = "red", size = 5) + # add text to plot
    labs(x='Number of Posts', y='Number of Followers', color='Channel Info', title='Scatterplot of Followers vs. Posts') # rename axes and add title

3.1.3 Bar Plots

To visualize the comparison of amounts

e.g. Out of the top 200 instagram accounts, which are the 2 most followed categories?

insta_cat <- insta |>
  group_by(Category) |>
  summarise(Total_Followers = sum(Followers))
insta_cat

A tibble: 12 × 2

Category	Total_Followers
<chr>	<dbl>
Beauty & Makeup	147400000
Craft/DIY	46200000
Finance	51900000
Health, Sports & Fitness	3438300000
Lifestyle	58100000
News & Politics	156000000
Not Available	984400000
entertainment	10859800000
fashion	511400000
food	93800000
photography	626100000
technology	122200000

ggplot(insta_cat, aes(x = Category, y = Total_Followers)) +
    geom_bar(stat = 'identity') +
    theme(text = element_text(size = 26))

ggplot(insta_cat, aes(x = Category, y = Total_Followers)) +
    geom_bar(stat = 'identity') +
    theme(text = element_text(size = 26))+
    coord_flip()

options(repr.plot.width = 12, repr.plot.height = 10)

# Reorder the levels of Category based on Total Followers

ggplot(insta_cat, aes(x = Total_Followers, y = reorder(Category, -Total_Followers))) +
    geom_bar(stat = 'identity', fill = "cornflowerblue") + 
    theme(text = element_text(size = 26)) + 
    labs(x = "Total Followers", y ="Category", title = "Barplot of Category by Total Followers ") # change labels and title

It is now clear that entertainment an Health, Sports and Fitness are the two most followed categories of Instagram accounts.

3.2 Rules of Thumb

Rule of Thumb: No tables / pie charts / 3D

Rule of Thumb: No tables / pie charts / 3D

Rule of Thumb: Use simple, colourblind-friendly colour palettes

• https://www.color-blindness.com/coblis-color-blindness-simulator/

Rule of Thumb: Include labels and legends, make them legible

Remember: a great visualization tells its own story without needing you to be there explaining

Rule of Thumb: avoid overplotting

Less is more!

diamond_plot <- ggplot(diamonds, aes(x = carat, y = price)) +
    geom_point() +
    xlab("Size (carat)") +
    ylab("Price (US dollars)")
diamond_plot

diamond_plot <- ggplot(diamonds, aes(x = carat, y = price)) +
    geom_point(alpha=0.2) + # alpha sets the transparency of points
    xlab("Size (carat)") +
    ylab("Price (US dollars)")
diamond_plot