CS 725/825 - Information Visualization
Fall 2013: Tues/Thurs 1:30-2:45pm, E&CS 2120

Print - Admin

Announcements

Staff

Schedule

Syllabus

Objectives

Project

Links


Tableau's data visualization software is provided through the Tableau for Teaching program.

InfoVis Lab

September 10-12, 2013

This lab will give you experience using Tableau, R, and d3 to create a variety of visualizations. The goal is to allow you to become familiar with these tools so that you are comfortable using them in your class assignments and final project.

Setup

Create a webpage to hold the results of your lab tasks.

Data

The data file that you'll use comes from http://www.sports-reference.com/cfb/years/2012-passing.html and lists the stats for FBS college quarterbacks in 2012.

I've also added ODU's Taylor Heinicke as the last row for comparison and indicated his conference as CAA (stats from http://stats.washingtonpost.com/cfb/players.asp?id=203530)

data file - http://www.cs.odu.edu/~mweigle/cs725f13/passing-stats.csv
metadata - http://www.cs.odu.edu/~mweigle/cs725f13/passing-stats.txt

If you have questions about the data, ask!

Tableau

We can use Tableau to get a quick look at the data, without the need for programming.

Use Tableau to create the following graphs:

  1. bar chart of passing yards per player (best displayed as a horizontal bar chart), with conference as color
  2. scatterplot of 2 interesting variables
  3. scatterplot of 2 interesting variables, split by conference
  4. one other interesting graph

For each graph, what insight does it provide? What other questions does it lead to?

R

R not only allows us to handle data and create graphs, but it also allows us to manipulate the data.

Use R to create the following graphs:

  1. scatterplot matrix of passing yards, passing TDs, passer rating, rushing yards, and rushing TDs
  2. use the gclus package to create a scatterplot matrix to color-code the scatterplots based on correlation
See http://www.statmethods.net/graphs/scatterplot.html. Your image should look something like this:
  1. bar chart of passing yards per player (best displayed as a horizontal bar chart), with conference as color
  2. bar chart of the average of one of the statistics for each conference

Did graph #2 help you gain any insights that you missed in graph #1?

rCharts can be used to create interactive R graphs. It provides a way to use several different JavaScript libraries to create the visualizations. They can all be exported to a webpage for sharing. The main github page at http://ramnathv.github.io/rCharts/ has examples of using several of these libraries.

Use rCharts to create the following interactive graphs and put the standalone interactive versions of each on your lab webpage:

  1. scatterplot of 2 interesting variables
  2. bar chart of the average of one of the statistics for each conference

You might want to look at these examples:

d3

d3 is our high-powered tool to create customized interactive visualizations.

Create the following graphs in d3:

  1. Modify the scatterplot you created for HW 1 so that it uses the football data. Plot two interesting variables.
  2. Modify the bar chart example at http://bl.ocks.org/weiglemc/6041345 to use the football data and add some interactivity.
Show the average of one of the statistics for each conference. You may modify the data file before you load it into d3, or you may do the calculations with JavaScript.

Other Toolkit or Library

Use any other toolkit or library that we've talked about or that you've found to create at least one graph using the football data. You might want to use http://selection.datavisualization.ch/ to help you choose.

HW 3

Anything that you don't complete in class on Thursday, finish as HW 3 by next Tuesday (Sep 17).

If you worked in a group in class, each member of the group should create a lab webpage and be responsible for understanding what was involved in creating the graphs.