Lecture #06: Exploratory

Data Analysis 2:
Graphical Data Analysis
 Menu
 Introduction
 Visualization of Data
 Graphical Presentation of Univariate Data

Bar Chart

Histogram

Dot Plots

Steam and Leaf Plots

Box and Whisker Plots

Pie and Donut Diagrams
 Graphical Presentation of Bivariate Data

Scatter Plots
 Graphical Presentation of Spatial Data

Contour Plots

Matrix Plots
 Other Types of Graphs

Star/Kite Plot

Chernoff Faces
 Concluding Remarks
 Visualization
Visualization is the conversion of data into a
visual or tabular format so that the
characteristics of the data and the
relationships among data items or attributes
can be analyzed or reported.
 Visualization of data is one of the most
powerful and appealing techniques for data
exploration.

Humans have a well developed ability to
analyze large amounts of information that is
presented visually

Can detect general patterns and trends

Can detect outliers and unusual patterns
 Example: Sea Surface Temperature
 The following shows the Sea Surface
Temperature (SST) for July 1982

Tens of thousands of data points are
summarized in a single figure
 Representation
 Is the mapping of information to a visual format
 Data objects, their attributes, and the relationships
among data objects are translated into graphical
elements such as points, lines, shapes, and
colors.
 Example:

Objects are often represented as points

Their attribute values can be represented as the
position of the points or the characteristics of the
points, e.g., color, size, and shape

If position is used, then the relationships of points, i.e.,
whether they form groups or a point is an outlier, is
easily perceived.
 Arrangement
 Is the placement of visual elements within a
display
 Can make a large difference in how easy it
is to understand the data
 Example:
 Selection
 Is the elimination or the de-emphasis of certain
objects and attributes
 Selection may involve the choosing a subset of
attributes

Dimensionality reduction is often used to reduce the
number of dimensions to two or three

Alternatively, pairs of attributes can be considered
 Selection may also involve choosing a subset of
objects

A region of the screen can only show so many points

Can sample, but want to preserve points in sparse
areas
 Goals of Graphing

1. Presentation of Descriptive Statistics

2. Presentation of Evidence
3. Some people understand subject
matter better with visual aids
4. Provide a sense of the underlying
data generating process (scatter-
plots)
 Which graph to use?
 Depends on type of data
 Depends on what you want to illustrate
 Depends on available statistical software
 Bar Chart
Birth Order of Spring 1998 Stat 250 Students

40

30
Percent

20

10

Middle Oldest Only Youngest

Birth Order
n=92 students
 Bar Chart
 Summarizes categorical data.
 Horizontal axis represents categories,
while vertical axis represents either counts
(“frequencies”) or percentages (“relative
frequencies”).
 Used to illustrate the differences in
percentages (or counts) between
categories.
 Visualization Techniques: Histograms
 Histogram

Usually shows the distribution of values of a single variable

Divide the values into bins and show a bar plot of the number of
objects in each bin.

The height of each bar indicates the number of objects

Shape of histogram depends on the number of bins
 Example: Petal Width (10 and 20 bins, respectively)
 Two-Dimensional Histograms
 Show the joint distribution of the values of two
attributes
 Example: petal width and petal length

What does this tell us?
 Histogram
Age of Spring 1998 Stat 250 Students

50
Frequency (Count)

40

30

20

10

18 19 20 21 22 23 24 25 26 27
Age (in years)
n=92 students
 Histogram
 Divide measurement up into equal-sized
categories.
 Determine number (or percentage) of
measurements falling into each category.
 Draw a bar for each category so bars’
heights represent number (or percent)
falling into the categories.
 Label and title appropriately.
 Too few categories
Age of Spring 1998 Stat 250 Students

60
Frequency (Count)

50

40

30

20

10

18 23 28
Age (in years)
n=92 students
 Too many categories
GPAs of Spring 1998 Stat 250 Students

6
Frequency (Count)

2 3 4
GPA
n=92 students
 Histograms
Histogram Histogram

25 25

20 20

15 15

10 10

5 5

0 0
.4 .8 .2 .6 48 .4 .8 .2 .6 65

.4

48

.4

65
.8

.2

.6

.8

.2

.6
34 37 41 44 to 51 54 58 61 to

34

51
37

41

44

54

58

61
to

to
to to to to .6 to to to to .6

to

to
to

to

to

to

to

to
.6

.6
31 .4 .8 .2 44 48 .4 .8 .2 61

31

44

48

61
.4

.8

.2

.4

.8

.2
34 37 41 51 54 58

34

37

41

51

54

58
Histogram

30

25
20

15

10
5

0
38

42

46

50

54

58

62

66

70

74
to

to

to

to

to

to

to

to

to

to
34

38

42

46

50

54

58

62

66

70
 Dot Plot

Fastest Ever Driving Speed

226 Stat 100 Students, Fall '98

100
Men

126
Women
70 80 90 100 110 120 130 140 150 160
Speed
 Dot Plot
 Summarizes measurement data.
 Horizontal axis represents measurement
scale.
 Plot one dot for each data point.
 Stem-and-Leaf Plot
Stem-and-leaf of Shoes N = 139 Leaf Unit = 1.0

12 0 223334444444
63 0 555555555555566666666677777778888888888888999999999
(33) 1 000000000000011112222233333333444
43 1 555555556667777888
25 2 0000000000023
12 2 5557
8 3 0023
4 3
4 4 00
2 4
2 5 0
1 5
1 6
1 6
1 7
1 7 5
 Stem and Leaf plots
Before:
N = 10 Median = 3.1 Quartiles = 3, 3.3
During:
N = 10 Median = 5.2 Quartiles = 2.3,
2 : 88 8.8
3 : 00112333
-1 : 0
After: -0 :
N = 10 Median = 5.5 Quartiles = 4.1, 6.7 0:
1:5
3 : 24
4 : 125 2 : 33
5: 3:3
6 : 567 4:
7:3 5:
6:
High: 17 7:1
8 : 68
9 : 15
 Stem-and-Leaf Plot
 Summarizes measurement data.
 Each data point is broken down into a
“stem” and a “leaf.”
 First, “stems” are aligned in a column.
 Then, “leaves” are attached to the stems.
Visualization Techniques: Box Plots
 Box Plots

Invented by J. Tukey

Another way of displaying the distribution of data

Following figure shows the basic part of a box plot

outlier

10th percentile

75th percentile

50th percentile
25th percentile

10th percentile
 Example of Box Plots
 Box plots can be used to compare
attributes
 Box Plot
 Summarizes measurement data.
 Vertical (or horizontal) axis represents
measurement scale.
 Lines in box represent the 25th percentile
(“first quartile”), the 50th percentile
(“median”), and the 75th percentile (“third
quartile”), respectively.
 An aside...
 Roughly speaking:

The “25th percentile” is the number such that
25% of the data points fall below the number.

The “median” or “50th percentile” is the
number such that half of the data points fall
below the number.

The “75th percentile” is the number such that
75% of the data points fall below the number.
 Box Plot (cont’d)
 “Whiskers” are drawn to the most extreme
data points that are not more than 1.5
times the length of the box beyond either
quartile.

Whiskers are useful for identifying outliers.
 “Outliers,” or extreme observations, are
denoted by asterisks.

Generally, data points falling beyond the
whiskers are considered outliers.
Using Box Plots to Compare
Fastest Ever Driving Speed
226 Stat 100 Students, Fall 1998
160
Fastest Speed (mph)

110

60
female male
Gender
 Pie Charts:
Proportions of Donut-Eating Professors by Weight Class

130-150
151-185
186-210
211-240
241-270
271-310
311+
Actually, why not use a donut
graph. Duh!
Proportions of Donut-Eating Professors by Weight Class

130-150
151-185
186-210
211-240
241-270
271-310
311+

See Excel for other options!!!!

 Which graph to use when?
 Stem-and-leaf plots and dotplots are good
for small data sets, while histograms and
box plots are good for large data sets.
 Boxplots and dotplots are good for
comparing two groups.
 Boxplots are good for identifying outliers.
 Histograms and boxplots are good for
identifying “shape” of data.
Visual Displays of Bivariate Data

Variable 1 Variable 2 Display

Example

Categorical Categorical Crosstabs

Categorical Continuous Box plots

Continuous Continuous Scatter plots

 Visualization Techniques:
Scatter Plots
 Scatter plots

Attributes values determine the position

Two-dimensional scatter plots most common,
but can have three-dimensional scatter plots

Often additional attributes can be displayed by
using the size, shape, and color of the markers
that represent the objects

It is useful to have arrays of scatter plots can
compactly summarize the relationships of
several pairs of attributes
• See example on the next slide
Scatter Plot Array of Iris
Attributes
With Outlier and Out of Range Value
M= 5.15,Sd= 3.67,Sk=-0.19,K=-1.51 r=-0.57, B=-0.6, t=-1.97, p=0.08, N=10

16
8

14
DURING,N=10
6

10 12
AFTER
4

8
2

6
0

4
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 0 2 4 6 8
Standard Normal Quantiles DURING
r=-0.57, B=-0.55, t=-1.97, p=0.08, N=10 M=6.35,Sd=3.82,Sk=2.01*,K=3.12*

16
8

14
6

AFTER,N=10
10 12
DURING
4

8
2

6
0

4 6 8 10 12 14 16 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

AFTER Standard Normal Quantiles
 Without Outlier
M= 5.15,Sd= 3.67,Sk=-0.19,K=-1.51 r=-0.92, B=-0.37, t=-6.33, p=0, N=9

7
8

6
DURING,N=10
6

AFTnew
4

5
2

4
0

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 0 2 4 6 8

Standard Normal Quantiles DURING
r=-0.92, B=-2.3, t=-6.33, p=0, N=9 M= 5.17,Sd= 1.50,Sk= 0.10,K=-1.67

7
8

6
6

AFTnew,N=9
DURING
4

5
2

4
0

4 5 6 7 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

AFTnew Standard Normal Quantiles
With Corrected Out of Range Value
M= 5.17,Sd= 1.50,Sk= 0.10,K=-1.67 r=-0.92, B=-2.09, t=-6.4, p=0, N=9
7

8
6
AFTnew,N=9

DURnew
6
5

4
4

2
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 4 5 6 7
Standard Normal Quantiles AFTnew
r=-0.92, B=-0.41, t=-6.4, p=0, N=9 M= 5.35,Sd= 3.37,Sk= 0.00,K=-1.81
7

8
DURnew,N=10
6
AFTnew

6
5

4
4

2 4 6 8 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

DURnew Standard Normal Quantiles
 Scatter Plots
Foot sizes of Spring 1998 Stat 250 students

31
30
Right foot (in cm)

29
28
27
26
25
24
23
22
22 23 24 25 26 27 28 29 30 31
Left foot (in cm)
n=88 students
 Scatter Plots
 Summarizes the relationship between two
measurement variables.
 Horizontal axis represents one variable
and vertical axis represents second
variable.
 Plot one point for each pair of
measurements.
 No relationship
Lengths of left forearms and head circumferences
of Spring 1998 Stat 250 Students
32
31
Left forearm (in cm)

30
29
28
27
26
25
24
23
22
52 57 62
Head circumference (in cm)
n=89 students
 Scatter-Plots – positive and
negative correlations
200.0 600.0
150.0 500.0
100.0 400.0
50.0 300.0
0.0 200.0
-50.0 0 20 40 60
100.0
-100.0 0.0
-150.0 0 20 40 60
 Visualization Techniques: Contour
Plots
 Contour plots

Useful when a continuous attribute is measured on a
spatial grid

They partition the plane into regions of similar values

The contour lines that form the boundaries of these
regions connect points with equal values

The most common example is contour maps of
elevation

Can also display temperature, rainfall, air pressure,
etc.
• An example for Sea Surface Temperature (SST) is provided
on the next slide
Contour Plot Example: SST Dec, 1998

Celsius
 Visualization Techniques: Matrix Plots
 Matrix plots

Can plot the data matrix

This can be useful when objects are sorted according
to class

Typically, the attributes are normalized to prevent one
attribute from dominating the plot

Plots of similarity or distance matrices can also be
useful for visualizing the relationships between objects

Examples of matrix plots are presented on the next two
slides
Visualization of the Iris Data Matrix

standard
deviation
Visualization of the Iris Correlation Matrix
 Visualization Techniques: Parallel
Coordinates
 Parallel Coordinates

Used to plot the attribute values of high-dimensional
data

Instead of using perpendicular axes, use a set of
parallel axes

The attribute values of each object are plotted as a
point on each corresponding coordinate axis and the
points are connected by a line

Thus, each object is represented as a line

Often, the lines representing a distinct class of objects
group together, at least for some attributes

Ordering of attributes is important in seeing such
groupings
Parallel Coordinates Plots
 Other Visualization Techniques
 Star Plots

Similar approach to parallel coordinates, but axes
radiate from a central point

The line connecting the values of an object is a
polygon
 Chernoff Faces

Approach created by Herman Chernoff

This approach associates each attribute with a
characteristic of a face

The values of each attribute determine the appearance
of the corresponding facial characteristic

Each object becomes a separate face

Relies on human’s ability to distinguish faces
Star Plots for Iris Data

Setosa

Versicolour

Virginica
Chernoff Faces
Setosa

Versicolour

Virginica
 Scales of Graphs
 It is very important to pay attention to the
scale that you are using when you are
plotting.
 Compare the following graphs created
from identical data
 Summary
 Examine all your variables thoroughly and
carefully before you begin analysis
 Use visual displays whenever possible
 Transform each variable as necessary to
deal with mistakes, outliers, and
distributions
 Concluding Remarks
 Many possible types of graphs.
 Use common sense in reading graphs.
 When creating graphs, don’t summarize
your data too much or too little.
 When creating graphs, label everything for
others.
 Remember you are trying to communicate
something to others!