Experimental Methods in Computer Science

Exercise 2 – How to Display Data

Goals

• Think about how to present data so that it tells the most convincing story and makes the most sense.
• Gain more experience with creating graphical representations of data.

Assignment

In ex1 we used simple graphs to learn about the Titanic data and see what is it's main story. Here the main point is precisely what to show and how to best show it.

Based on the ideas derived in Ex1, we want to show the following aspects of the data:

• The relative number of people in different classes and crew
• The different survival rate as a function of class (including crew)
• The different survival rate at a function of gender

The challenge is to create a single graph that shows as much as possible, highlights the most important point(s), and at the same time is easy to interpret and looks good. The most important characteristic of a good graph is that it tells the story loud and clear. If you need to explain it, it is not clear enough.

In designing the graph, you should focus on what you want the graph to show. Choose the representation that seems best for this need. It may be a good idea to experiment with several different designs. This way you can check what works best.

Of course, this includes labeling the axes, providing a legend or annotations, etc.

Submit

Use moodle to submit a report on your work, in pdf. The report should include

1. Your names, logins, and IDs
2. A short explanation of what you did, including
   1. What was the main goal? If you decided not to include something, why not?
   2. What graphical representation did you choose? How does it support the above goal?
   3. If you tried more than one design, what were they and why did you chose one over the other?
3. The output plot.

Submission deadline is Monday, 28/2/11, so I can give feedback in class on Tuesday.

Please do the exercise in pairs.

To the course home page