ETX2250/ETF5922: Data Visualization and Analytics
Introduction to visualization
Lecturer:
Department of Econometrics and Business Statistics
Week 1
When have you seen visualizations used effectively?
A history lesson: Cholera
In 1854 there was a breakout of the cholera disease in London killing 616 people At the time it was speculated that the disease was carried in the air
A physician called was sceptical and began to collect data…
3/36
4/36
5/36
Consequences
The map showed the cholera was more prevalent around a water pump on Broad Street The pump was closed down.
Eventually it was established that cholera is a water-borne disease.
Data visualization saves lives!
6/36
Crimean War
At the same time Great Britain was at war against Russia in the Crimean peninsula
is famous as a nurse who treated the wounded soldiers
She also advocated to the British Parliament for more sanitary conditions in military hospitals. She knew the power of using data visualization
7/36
8/36
Nightingale’s Rose chart
Blue areas: Preventable deaths
Red areas: Deaths from battle wounds Black areas: Other causes
9/36
Aftermath
The improved sanitation at military hospitals was eventually implemented in civilian hospitals. Data visualisation saves lives.
became the rst female member of the Royal Statistical Society.
10/36
1812 Napoleon thought it was a good idea to invade Russia.
This campaign was a disaster for the French.
Engineer Charles captured the extent of this catastrophe using visualisation.
11/36
12/36
Minard’s plot
This visualisation provides information on 6 variables in one chart. Number of troops.
Whether troops advance or retreat.
Temperature and time.
Longitude and latitude.
Despite the clear message that invading Russia in winter is a bad idea, some people did not learn this lesson.
13/36
A bit more on Minard
14/36
Minard was so well regarded that all Ministers of Public Works in France had his work in the background of their portraits from 1850-1860
Source: https://infowetrust.com/ nding-minard
15/36
Why data visualization?
Gain insights into information from data by mapping it to graphical elements
Provide qualitative overview of large data sets
Search for patterns, trends, structure, irregularities, relationships among data
It helps to nd interesting regions and suitable parameter for further quantitative analysis. It is the starting point of every Business Analytics Project
16/36
Types of visualization
class: transition
Bar plot
Display of tabulated frequencies, shown as bars The height of the bar is the important factor
18/36
Scatter Plot
Provides a rst look of the relationship between 2 variables Possible to see clusters of similar grouped points
Each pair of values is treated as a pair of coordinates and plotted
19/36
Boxplot
Summary of a distribution
Minimum, Q1, Median, Q3, Maximum
Compare the distribution of a numerical variable across categories
20/36
Steps in a Data Project
Problem Speci cation Data Visualisation
Data Analysis
Evaluation & Interpretation Communication
And re-iterate over: Ideally, automate or ensure reproducibility at as many steps as possible
21/36
Programming
class: transition
Programming and visualization
It allows us to have reproducible steps, which can be applied for many different data sets
Make sure the analysis is not just point and click, you can work as a team on it on the same code Same goes for analysis
23/36
R: A programming language
It is like learning a new language (so you need to think about learning the grammar and structure, and how to communicate well in it!)
It has been around for a while
It is regularly maintained and is open source
Below you see code that plots points on a plane, colours it by country and adds the text.
P <- ggplot (data = protein.df,
aes (x = RedMeat,
y = WhiteMeat,
colour = Location,
label = Country)) +
geom_point () +
geom_text (size = 3,
p
check_overlap = TRUE)
24/36
Installing R
First install R
Got to https://www.r-project.org/
Click "download R"
Select a mirror (I use the Melbourne one) Install for your operating system
Then install R studio (an IDE that we use to work with R)
Got to https://www.rstudio.com/products/rstudio/download/ (you only need the free version) Select download for your system
Follow the prompts to install
Always interact with R using Rstudio
25/36
Working in projects
One challenge with working with code is having to manage your working directory and le paths (to tell the computer where things are located)
To make this easier, we will use projects
For the rst tutorial, you will download a zipped folder from Moodle. Inside you will nd the following les:
etx2250-etf5922-lecture-01
├── analysis
│ └── tutorial_week01.Rmd
└── etx2250-etf5922-lecture-01.Rproj
Make sure you open by double clicking the .Rproj or by using File>Open Project in Rstudio
26/36
Glossary and important concepts
Assignment of a variable
x <- 5
Functions are like recipes. They take ingredients, perform a set of steps, and then return a result.
result <- function_name(input1 = , input2 = ,... )
In R the format of a function starts with a function_name, then specify any input needed, like data
27/36
Glossary and important concepts
Packages are collections of functions. You need to download them using the following code. Try running this command before your tutorial in the bottom left hand panel of your R studio.
install.packages(c("ggplot2", "tidyverse"))
Once you've installed them, to use them in a speci c R studio session use the library command
library(ggplot2) library(tidyverse)
28/36
Glossary and important concepts
Data is stored in different forms in R. Two we will use are: Variable
Data Frame, has rows and columns. Each column is a named variable Data has different types in R. Each variable must be made of the same type:
numeric (numbers)
integers (only round numbers)
factors (a set list of possible values. Stored as labels, but also have underlying numbers) characters (text data)
29/36
Grammar of graphics
“
A grammar of graphics is a tool that enables us to concisely describe the components of a graphic. Such a grammar allows us to move beyond named graphics (e.g., the ``scatterplot'') and gain insight into the deep structure that underlies statistical graphics.
ggplot(data = ) +
mapping = aes(
stat =
postion =
)+
)
. A layered grammar of graphics. Journal of Computational and Graphical Statistics, vol. 19, no.
1, pp. 3–28, 2010.
30/36
Steps for plotting
Step 1. How will you obtain the data?
Commonly in this course: read from a csv (data values separated by a common)
Step 2. What do you want in the graph?
What information are you interested in?
What question are you trying to answer?
Who do you want to share what you have learned with?
31/36
Steps for plotting
Step 3. How do you want to represent your information? Scatter plot vs bar graph?
How many variables are you dealing with?
What comparisons do you want to consider?
Step 4. Figure out what you want to include in the aesthetics of your graph? x,y coordinates (which variable is plotted on each)
Could you add colour to create new insights? Are the colours accessible? Labels – is the plot readable and easy for others to understand?
32/36
Data manipulation
The data that you will get will never be in the form you can use it directly A nice tool in R is the pipe operator: “%>%”
It allows you to pass the information in a data frame from the left to the right, to the next command. Like a stream.
It will make more sense in the tutorial when you get to try things out
33/36
Common issues at the beginning
Close parenthesis for every function you open a bracket ( there also needs to be a closing one ) Check parenthesis again if you are getting strange errors
Syntax needs time to get used too, everyone struggles at the beginning.
Take notes of what each function does, and when you would use it. It’s like learning the new words of a new language
34/36
Live coding example
class: transition
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Lecturer:
Department of Econometrics and Business Statistics
Week 1