DCF: Data and Computing Fundamentals at Macalester

• Provide a quick foundation in the techniques covered in DCF
• Inspire topics for data-oriented modules in your own courses
• Solicit your input for revisions to DCF
• Start the process of collaboration to develop modules and revise DCF

Provide meaningful data processing display and analysis skills

• do productive work
• applicable to diverse disciplines
• forward looking: seriously multivariable

Fits in early, with minimal alterations to a student's schedule *Time budget: 10 class hours, 20 out-of-class hours

No pre-requisites

Fun, not a slog

1. Graphical Displays

2. Workflow

3. Relational Operations

4. Transformations

5. Modeling

Class organized around case studies data settings.

Wednesday

• Examples of Data settings & Graphical displays
• Skills: RStudio, R patterns, Workflow

Thursday

• Relational operations, data transformations
• Skills: Making graphics

Friday

• Modeling
  • Recent example of modeling using big data: American Revolutionaries

By a data setting, we mean:

• One or more related data files
• At least one data organization, processing, modeling or visualization example.
• A set of questions for the students to investigate using the data.

Your task: Think about a data setting that you would like to have in your course(s). Meanwhile, we'll show some examples …

We chose some data sets for the prototype course to illustrate various aspects of data organization and display.

Criteria:

• Potentially of interest to students
• Topics generally accessible: biologists, chemists, economists, physicists, etc.
• Illustrates an important aspect of data organization, data processing, modeling, and/or data visualization

• Data from the UN Food and Agriculture Organization
• Gapminder country-by-country
• Bird observations at Ordway
• NHANES 1999-2004 health-outcomes
• Centers for Medicare Hospital cost comparison
• Voters in Wake County, North Carolina

Birds captured and released at the Katharine Ordway Natural History Study Area, a 278-acre preserve in Inver Grove Heights, Minnesota, owned and managed by Macalester College. Source: Jerald Dosch, the manager of the Study Area. Currently 15829 cases in 23 variables.

Error: object 'OrdwayBirdsOrig' not found

• Data cleaning
  • Species names are not-standardized
  • Inconsistent formatting of quantitative data
• Patterns in season and time-of-day
  • Discover what birds are migratory
• Discovering classification criteria

Age, sex, height, wgt, etc. along with

• cholesterol, systolic blood pressure
• BMI, and other derived measures
• smoking status, diabetic, mortality,

31126 people.

• Simple depictions of relationships, e.g. height and age
• Comparison of models — Can we see the difference between girls' and boys' growth curves.
• Modeling of risk for diabetes

500,000 registered voters from Wake County North Carolina.

Use as a Data Setting

• Orientation to tabulation in a familiar context
• Simple transformations (e.g. age groups)

We chose what we think are the major categories of accessible visualization modalities:

• scatter plots
• bar charts
• maps
• parallel coordinates
• trees

Your task: Think about what modalities are important to you.

• What's the story told by this graph?
• How would you need to reorganize the data?
• What other display choices would be informative?
• Does this graph inform the story you want to tell?
• If not, what graph would you want to make?

We all have preferred ways of working:

• How we store and arrange files.
• How we write up reports.
• How we keep track of what we do.

Sophisticated use of the computer is difficult.