Department of Statistics and Biostatistics
California State University, East Bay
Winter 2015
|
|
Statistics 4970/6610: Data Visualization
Week 11:
Here are some other things that we have not gotten to in the class.
Website Spotlight: datavis.ca
Website Spotlight:
Edward Tufte
Website Spotlight: ddpy
Thank you too the following former students.
1. I wanted to talk about social media and networks. We did not get
there. Try gephi This is an
open source software package for the visualization of networks.
Check out SoMe Lab. And
Jeff Hemsley's website.
Thank you Jeff for showing me some of what you are interested in.
2. Rladies
Gabriela de Queiroz
Gabriela introduced me to FlowingData and she encouraged me to move forward with
learning more about Visualization. Everyone should join Rladies!
3. Analytics Hackathon
Charles
Twitchell Charles introduced me to more of the Bay Area MeetUps.
His interested in finance has lead me to the use of many more R libraries for
finance and has lead me to learning python.
Networks
SAS and Teradata.
Turns out that students can use SAS on basically any platform now with
SAS
University. I would suggest everyone give it a try. And
SAS Visual Analytics software is now part of the
TeraData University.
python
Some further things that I wanted to discuss related to python.
Next Quarter: APIs and cloud computing for Machine Learning
Week 10:
- Final next week Wednesday 3/18/2015 8:00-9:50
(Correction, the final starts at 8:00pm.)
- Activity Two: (Optional. This is for those interested
in python.)
- Using python
- Try my updated version of the authors python code. My
code does not currently work as well since it is not using
google's API
geocode-locations3.py
- Activity Three:
- Using R
- Try the R code for the first part of Ch. 8
mapping.R
Last Day: Going Foward
Software
More D3 and js
Microsoft tools for Visualization
SAS tools for Visualization
SAS University
blogs
Further Education
Week 9:
- Homework 8 has been posted. Note that an additional
Activity has been added, Activity Five. See below
- Project 2: Pick your favorite plot from the book.
Find a dataset that you can used to make such a plot.
Produce the plot. Explain what the data and variables
represent. Explain how the graph displays a visualization
of the information in the data.
- Activity Three:
- Download the data from the
Book
website for Chapter 7.
- Make the Star Charts and Nightingale Charts for the
Crime data.
- Activity Four:
- Download the data from the
Book
website for Chapter 7.
- Make the Parallel Coordinates Charts for the Education data.
- The libary ggparallel can be used in R to make parallel
coordinate plots.
parallel_examples.R
- Software Spotlight -
ggobi
- Activity Five:
- Download the data from the
Book
website for Chapter 7.
- Make the plots for the Education data using Multidimensional
Scaling and using Clustering.
Week 8:
- Homework 7 has been posted.
- Activity One:
- From the
Correlation presentation, from last week, work with the book
price data to show the effect of Simpson's Paradox. Use
Minitab to fit the model. Use tableau to fit the model
- Do you see how to plot the individual lines on the
scatterplot. Do you see how to fit the model?
- Activity Two:
- Download the data from the
Book
website for Chapter 6.
- Make the histogram for the birth-rate data.
- Try the option, brakes=5 and brakes=20
- Some interesting papers by Hadley Wickham
- Website Spotlight -
iPlots
- Activity Five:
- Try the code in plot-tv-sizes.R
Note the use of the par() function in R.
- Activity Six:
- Try the hts01.R example code.
Week 7:
- Correlation
- We will use the Chapter 6 data from the
Book
website, at the bottom of the page.
- From the Quiz:
- Quiz 1 solution has been posted on Blackboard under Course
Materials.
- Midterm next time in class.
- Homework 6 due date is now Monday February 23.
- Project 1 will be assigned on Wednesday.
- Class Spotlight -
UW INFX 598
- Website Spotlight -
Canvas
- Activity One:
- From today's presentation
Correlation
try the code for the Simpson's Paradox.
- Do you see how the direction of the correlation changes with
the confounding variable?
- Activity Two:
- Download the data from the
Book
website for Chapter 6.
- Make the scatterplot matrix for the crime data.
- Activity Three:
- Make the correlogram for the crime data.
- Activity Four:
- Try to recreate the dataset needed to make the plot from the
Wall Street Journal graph, Chinese Labor Rifts Deepen by using
the website
Chinese Labor Bulletin
Strike Map
- Be sure to use area in your Bubble plot.
- What do you see that might be problem from the website?
Week 6:
- I have figured out how to create an Assignment in Blackboard
under Course Materials. Please try to turn in your
Homework 5 in Blackboard.
- Stat. 6610 Extra-Credit
- Proportions2
- Quiz 1 next week Wednesday Feb. 11 The quiz will cover
the Chapters 1 - 5 in the book.
- Today we will be in the computer lab.
- We will use the Chapter 5 data from the
Book
website, at the bottom of the page.
- FlowingData
-
Tutorial D3.js
- FlowingData
-
Tutorial Treemap
- Activity One:
Protovis
- Open
donut.html Right-click and Inspect Element.
Click on Sources and download the donut.html (double click on
the file name) and protoviz-r3.2.js files to your computer.
In the directory where you have downloaded donut.html, make a
directory js, and save protoviz-r3.2.js to that directory.
- Open donut.html in a text editor such as
NoteTab Light. You
will need to remove the / before the js/protovis-r3.2.js
- Replace the data in the donut.html file with the following
data
- var data =
[200,150,100,90,80,70,50,20,10,1];
- And find the color line:
- var depthColors = pv.Scale.linear(0, 200)
- Activity Two:
Protovis
- Open
stacked-bar.html
- Click on Sources and download the stacked-bar.html (double
click on the file name) and protoviz-r3.2.js files to your
computer. In the directory where you have downloaded
stacked-bar.html, make a directory js, and save protoviz-r3.2.js
to that directory.
- Open stacked-bar.html in a text editor such as
NoteTab Light. You
will need to remove the / before the js/protovis-r3.2.js
- Change the first bar in the stacked-bar.html to 50, 45, 5
- Change the title of the graph to
- <h1>Approval Ratings for President Obama</h1>
- Go to the 0to225.com
and pick 3 new colors and implement the change.
- Change the labels from "white" to "black".
- Activity Five: tableau
- Take the data from the
donut.html
example and enter it into tableau.
donut.xlsx
- var data = [172,136,135,101,80,68,50,29,19,41];
var cats = ["Statistics", "Design", "Business", "Cartography",
"Information Science", "Web Analytics", "Programming",
"Engineering", "Mathematics", "Other"];
- Make a pie chart.
- What do the colors looks like?
- Change the color palette to Orange.
- Change the transparency to 50%.
- In tableau, create Story and add a title and a lead in
sentence.
- An example, donut.twb
- Activity Six: tableau
- Make a Treemap for the
post-data.txt that looks as close to the picture in the book
on page 161.
- Activity Seven: tableau
- Make time plots for the
expenditures.txt data that looks as close to the picture in
the book on page 172.
- Hint: If you ever run into
problems running your R code it might be due to some previous
things being in your workspace. Try removing all of the
save objects with rm( list = ls() ) (Thank you Emily.)
Week 5:
- TimeSeries2
- Today we will be in the computer lab.
- We will discuss the first Project, the Quiz, and Midterm
this week. (Current plan is for a short quiz next week on
Wednesday. The Midterm the following Wednesday. The
first project given out next week.)
- We will use the Chapter 4 data from the
Book
website, at the bottom of the page.
- Activity One: Hot Dog Eating - bar graphs
- Activity Two: FlowingData subscribers - time plots
- Activity Three: World Population - time plots
- Activity Four: US Postage - step plot
- Activity Five: Unemployment Data - LOESS
- Activity Six: Johnson & Johnson stock price - decompose( )
function in R, ACF, PACF
Week 4:
- Tools2
- Today we will be in the computer lab.
- Activity Four: See D3 in action
Week 3:
Week 2:
- Monday will be the first day of class that we will be
meeting in the computer lab SSc 146, not in the classroom.
- All twitter suggestions are optional.
- DataScraping
- Spotlight Website -
Data Driven
Journalism
- Spotlight Website -
Automatic (Thank
you Josh. This is so cool!!!)
Week 1:
- Getting Started
- Spotlight Software -
Gapmider World Offline
- My time series plot of Chevron stock prices. Made with tableau
cvx
- Spotlight Software -
tableau public
- To find the videos about how to get started with tableau, go to the
tableau public website and click on the top link HOW IT WORKS. There
is a nice video on this page. To the right the fourth link is to
Training. This link takes you to many video about how to use tableau.
Please note the
online training this Friday at 9:30 AM. Hope you can find time to
register and attend this online training.
- For the first day of class we will be meeting in NSc 207, not in the
computer lab.
- Welcome
- Spotlight Software - clic<tale
|