Data Methods


Timeline

Show & Tell + Discussion 15 minutes
Data Science Methods 40 minutes
Methods Discussion 20 minutes
Data Analysis and Lobby Design 1 hour
Reflection and Presentation 15 minutes
Total 150 minutes

Dataset

You will be using the data you collected as part of the mini project. If you don’t already, have the raw data file prepared to use.

Repository

https://github.com/CSCI-4830-002-2014/hackathon-datamethods

Please fork the README file from this repository and use it to submit all your responses.

Objectives

  1. Begin developing a method for approaching data science questions
  2. Do a preliminary analysis of the data collected in the mini-project
  3. Start coming up with design ideas for the final project

Prerequisites

Teams

For the first part of class when discussing data methods, you will work in small groups of 2-3 to answer these questions.

During the second half of class, we will divide into two groups, sending half to the lobby and leaving half of the class here in the lounge. You can continue working in the same groups, but I want each person to work with their own data for the analysis section.

Objective 1: Data Science Methods

For this section, you will answer the 15 questions listed in Part 1 of the submission document. Don’t worry if you aren’t sure about some of the questions, we’ll go over them quickly after.

Some of the questions will only take a minute to answer and others may need some thought.

Objective 2: Data Analysis

Use Part 2 of the submssion document for this objective.

For this section we will use gauss along with the data we collected in the mini-project to do some basic analysis and try to figure out what we can visualize. You can also use JavaScript’s built-in Math methods.

Use callback functions to your advantage in order to get your results clean for submission.

You will need to submit the following results:

  1. Screenshot proving the data was successfully loaded into Gauss
  2. What is the most frequent value in your data
  3. What is the range of your data
  4. Identify when the biggest change in your data occurred
  5. What shape might your data take if graphed by value on the x-axis and frequency on the y-axis? [hint: use .distribution(‘relative’)]
  6. What value threshold represents the signal you were looking for in your experiment?
  7. What percentage of your data falls above or below that threshold?
  8. Is the signal you’re looking for in the data? Why or why not? Justify your answer with some analysis.

Objective 3: Final Project Design

Use Part 3 of the submssion document for this objective.

For this section, you will go down into the ATLAS lobby to answer some questions and brainstorm what could be possible for our final class project.

Remember our class project will be an installation in this space that uses the main screen in the entranceway of the lobby and, secondarily, the projectors near the stairwell. You can do anything you want as long as it involves collecting data. Be creative!

You’ll need submissions for the following questions:

  1. Link to a sensor(s) or interactive device(s) that you would like to use. Feel free to use Sparkfun or Instructables for ideas.
  2. What would it measure and how?
  3. Where would you put it?
  4. What problems could threaten the validity of your data?
  5. How often would you take a sample and when would you dump it to the database?
  6. How might you visualize the final result?
  7. Would this exhibit always result in interesting data or only sometimes? How might you trigger it to only collect at the right time? (if this doesn’t apply to your idea, specify why in your response)

Presentation

There will be 2 presentation phases. The first one will be after the data methods meaning you’ll need to push your responses up before the discussion begins so we can discuss them as a class.

The second will be after the second half of class where we will see what results people got from their data and what designs ideas everyone had.