ETC1010: Data Modelling and ComputingWeek of Data Visualisation: Lecture 3Dr. Nicholas Tierney & Professor Di CookEBS, Monash U.2019-08-141 / 46

Learning Tips2 / 46

Understanding learning

Growth and fixed mindsets
Reframe success + failure as opportunities for growth
Growing area of research by Carol Dweck of Stanford

3 / 46

Reframing

From

"I'll never understand"

"I just don't get programming"

"I'm not a maths person"

4 / 46

Reframing

From

"I'll never understand"

"I just don't get programming"

"I'm not a maths person"

To

"I understand more than I did yesterday"

"I can learn how to program"

"Compared to this last week, I've learnt quite a bit!"

4 / 46

Overview for todayGoing from tidy data to a data plot, using a grammar
Mapping of variables from the data to graphical elements
Using different geoms

5 / 46

Example: Tuberculosis data

The case notifications table From WHO.

Data is tidied here, with only counts for Australia.

tb_au

## # A tibble: 192 x 6
##    country   iso3   year count gender age  
##    <chr>     <chr> <dbl> <dbl> <chr>  <chr>
##  1 Australia AUS    1997     8 m      1524 
##  2 Australia AUS    1998    11 m      1524 
##  3 Australia AUS    1999    13 m      1524 
##  4 Australia AUS    2000    16 m      1524 
##  5 Australia AUS    2001    23 m      1524 
##  6 Australia AUS    2002    15 m      1524 
##  7 Australia AUS    2003    14 m      1524 
##  8 Australia AUS    2004    18 m      1524 
##  9 Australia AUS    2005    32 m      1524 
## 10 Australia AUS    2006    33 m      1524 
## # … with 182 more rows

6 / 46

The "100% charts"

ggplot(tb_au, aes(x = year, y = count, fill = gender)) +
  geom_bar(stat = "identity", position = "fill") +
  facet_grid(~ age) +
  scale_fill_brewer(palette="Dark2")

7 / 46

100% charts, is what excel names these beasts. What do we learn?

Let's unpack a bit.8 / 46

Data Visualisation

"The simple graph has brought more information to the data analyst’s mind than any other device." — John Tukey

9 / 46

Data VisualisationThe creation and study of the visual representation of data.
Many tools for visualizing data (R is one of them)
Many approaches/systems within R for making data visualizations (ggplot2 is one of them, and that's what we're going to use).

10 / 46

ggplot2 $\in$ tidyverse

ggplot2 is tidyverse's data visualization package
The gg in "ggplot2" stands for Grammar of Graphics
It is inspired by the book Grammar of Graphics by Leland Wilkinson ^†
A grammar of graphics is a tool that enables us to concisely describe the components of a graphic

^† Source: BloggoType

11 / 46

From BloggoType

12 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au)

13 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year,
           y = count))

14 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year, 
           y = count))  +
  geom_point()

15 / 46

Our first ggplot! (what's the data again?)
 
    country 
    iso3 
    year 
    count 
    gender 
    age 
  


    Australia 
    AUS 
    1997 
    8 
    m 
    1524 
  

    Australia 
    AUS 
    1998 
    11 
    m 
    1524 
  

    Australia 
    AUS 
    1999 
    13 
    m 
    1524 
  

    Australia 
    AUS 
    2000 
    16 
    m 
    1524 
  

    Australia 
    AUS 
    2001 
    23 
    m 
    1524 
  

    Australia 
    AUS 
    2002 
    15 
    m 
    1524 
  

    Australia 
    AUS 
    2003 
    14 
    m 
    1524 
  

    Australia 
    AUS 
    2004 
    18 
    m 
    1524 
  

    Australia 
    AUS 
    2005 
    32 
    m 
    1524 
  

    Australia 
    AUS 
    2006 
    33 
    m 
    1524 
  


16 / 46

country	iso3	year	count	gender	age
Australia	AUS	1997	8	m	1524
Australia	AUS	1998	11	m	1524
Australia	AUS	1999	13	m	1524
Australia	AUS	2000	16	m	1524
Australia	AUS	2001	23	m	1524
Australia	AUS	2002	15	m	1524
Australia	AUS	2003	14	m	1524
Australia	AUS	2004	18	m	1524
Australia	AUS	2005	32	m	1524
Australia	AUS	2006	33	m	1524

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year, 
           y = count))  +
  geom_col()

17 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year, 
           y = count,
           fill = gender))  +
  geom_col()

18 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year, 
           y = count,
           fill = gender))  + 
  geom_col(position = "fill")

19 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year, 
           y = count,
           fill = gender))  + 
  geom_col(position = "fill") +
  scale_fill_brewer(
    palette = "Dark2"
    )

20 / 46

Our first ggplot!

library(ggplot2)
ggplot(tb_au,
       aes(x = year, 
           y = count,
           fill = gender))  + 
  geom_col(position = "fill") +
  scale_fill_brewer( 
    palette = "Dark2"
    ) +
  facet_wrap(~ age)

21 / 46

First argument provided is the name of the data, tb_au
Variable mapping: year is mapped to x, count is mapped to y, gender is mapped to colour, and age is used to subset the data and make separate plots
The column geom is used, geom_col
We are mostly interested in proportions between gender, over years, separately by age. The position = "fill" option in geom_bar sets the heights of the bars to be all at 100%. It ignores counts, and emphasizes the proportion of males and females.

The "100% charts"

ggplot(tb_au, aes(x = year, y = count, fill = gender)) +
  geom_bar(stat = "identity", position = "fill") +
  facet_grid(~ age) +
  scale_fill_brewer(palette="Dark2")

22 / 46

The "100% charts"

ggplot(tb_au, aes(x = year, y = count, fill = gender)) +
  geom_bar(stat = "identity", position = "fill") +
  facet_grid(~ age) +
  scale_fill_brewer(palette="Dark2")

What do we learn

22 / 46

100% charts, is what excel names these beasts. What do we learn?

What do we learn?Focus is on proportion in each category. 
Across (almost) all ages, and years, the proportion of males having TB is higher than females
These proportions tend to be higher in the older age groups, for all years.

23 / 46

Code structure of ggplotggplot() is the main function
Plots are constructed in layers
Structure of code for plots can often be summarised as

ggplot(data = [dataset], 
       mapping = aes(x = [x-variable], 
                     y = [y-variable])) +
   geom_xxx() +
   other options

24 / 46

How to use ggplot

To use ggplot2 functions, first load tidyverse

library(tidyverse)

For help with the ggplot2, see ggplot2.tidyverse.org

25 / 46

Let's look at some more options to emphasise different features

26 / 46

ggplot(tb_au,
       aes(x = year, 
           y = count,
           fill = gender))  + 
  geom_col(position = "fill") +
  scale_fill_brewer( 
    palette = "Dark2"
    ) +
  facet_wrap(~ age)

27 / 46

Emphasizing different features with ggplot2

ggplot(tb_au, 
       aes(x = year, 
           y = count, 
           fill = gender))  + 
  geom_col(position = "fill") +
  scale_fill_brewer( palette = "Dark2") +
  facet_grid(~ age)

28 / 46

Emphasise ... ?

ggplot(tb_au, 
       aes(x = year, 
           y = count, 
           fill = gender))  + 
  geom_col() +
  scale_fill_brewer( palette = "Dark2") +
  facet_grid(~ age)

29 / 46

What do we learn?, position = "fill" was removed
Focus is on counts in each category. 
Different across ages, and years, counts tend to be lower in middle age (45-64)
1999 saw a bit of an outbreak, in most age groups, with numbers doubling or tripling other years.
Incidence has been increasing among younger age groups in recent years.

30 / 46

Emphasise ... ?

ggplot(tb_au, 
       aes(x = year, 
           y = count, 
           fill = gender)) +
  geom_col(position = "dodge") +
  scale_fill_brewer(palette = "Dark2") +
  facet_grid(~ age)

31 / 46

What do we learn?, position="dodge" is used in geom_col
Focus is on counts by gender, predominantly male incidence.
Incidence among males relative to females is from middle age on. 
There is similar incidence between males and females in younger age groups.

32 / 46

Separate bar charts

ggplot(tb_au, 
       aes(x = year, y = count, fill = gender)) +
  geom_col() +
  scale_fill_brewer(palette = "Dark2") +
  facet_grid(gender ~ age)

33 / 46

What do we learn?facet_grid(gender ~ age) + faceted by gender as well as age
note facet_grid vs facet_wrap
Easier to focus separately on males and females.
1999 outbreak mostly affected males.
Growing incidence in the 25-34 age group is still affecting females but seems to be have stablised for males.

34 / 46

Pie charts? Rose Charts

ggplot(tb_au, 
       aes(x = year, y = count, fill = gender)) +
  geom_col() +
  scale_fill_brewer(palette="Dark2") + 
  facet_grid(gender ~ age) +
  coord_polar() +
  theme(axis.text = element_blank())

35 / 46

What do we learn?Bar charts in polar coordinates produce rose charts. 
coord_polar() + plot is made in polar coordinates, rather than the default Cartesian coordinates
Emphasizes the middle years as low incidence. 

36 / 46

Rainbow charts?

ggplot(tb_au, aes(x = 1, 
                  y = count, 
                  fill = factor(year))) +
  geom_col(position = "fill") +
  facet_grid(gender ~ age)

37 / 46

What do we see in the code??A single stacked bar, in each facet. 
Year is mapped to colour. 
Notice how the mappings are different. A single number is mapped to x, that makes a single stacked bar chart.
year is now mapped to colour (that's what gives us the rainbow charts!)

38 / 46

What do we learn?Pretty chart but not easy to interpret. 

39 / 46

(Actual) Pie charts

ggplot(tb_au, aes(x = 1, y = count, fill = factor(year))) +
  geom_col(position = "fill") +
  facet_grid(gender ~ age) +
  coord_polar(theta = "y") +
  theme(axis.text = element_blank())

40 / 46

What is different in the code?coord_polar(theta="y")  is using the y variable to do the angles for the polar coordinates to give a pie chart.

41 / 46

What do we learn?Pretty chart but not easy to interpret, or make comparisons across age groups. 

42 / 46

Why?

The various looks of David Bowie

Using named plots, eg pie chart, bar chart, scatterplot, is like seeing animals in the zoo.
The grammar of graphics allows you to define the mapping between variables in the data, with elements of the plot.
It allows us to see and understand how plots are similar or different.
And you can see how variations in the definition create variations in the plot.

43 / 46

Your Turn:Do the lab exercises
Take the lab quiz
Use the rest of the lab time to coordinate with your group on the first assignment.

44 / 46

References

45 / 46

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

↑, ←, Pg Up, k	Go to previous slide
↓, →, Pg Dn, Space, j	Go to next slide
Home	Go to first slide
End	Go to last slide
Number + Return	Go to specific slide
b / m / f	Toggle blackout / mirrored / fullscreen mode
c	Clone slideshow
p	Toggle presenter mode
t	Restart the presentation timer
?, h	Toggle this help

ETC1010: Data Modelling and Computing

Week of Data Visualisation: Lecture 3

Dr. Nicholas Tierney & Professor Di Cook

EBS, Monash U.

2019-08-14

Learning Tips

Understanding learning

Reframing

From

Reframing

From

To

Overview for today

Example: Tuberculosis data

The "100% charts"

Let's unpack a bit.

Data Visualisation

Data Visualisation

ggplot2 ∈∈ tidyverse

Our first ggplot!

Our first ggplot!

Our first ggplot!

Our first ggplot! (what's the data again?)

Our first ggplot!

Our first ggplot!

Our first ggplot!

Our first ggplot!

Our first ggplot!

The "100% charts"

The "100% charts"

What do we learn?

Code structure of ggplot

How to use ggplot

Emphasizing different features with ggplot2

Emphasise ... ?

What do we learn?

Emphasise ... ?

What do we learn?

Separate bar charts

What do we learn?

Pie charts? Rose Charts

What do we learn?

Rainbow charts?

What do we see in the code??

What do we learn?

(Actual) Pie charts

What is different in the code?

What do we learn?

Why?

Your Turn:

References

Share and share alike

Learning Tips

Help

ggplot2 $\in$ tidyverse