D3.js & Crossfilter: Explore Toronto Fire Trends

VISUALIZATION: A live-coding session where you will learn how to use D3.js and Crossfilter. This event is for Data Heads who are looking for new tools to explore and visualize large datasets. Some background in javascript is helpful.

We will be using a combination of D3.js (http://bit.ly/ViygYh) and Crossfilter (http://bit.ly/VYuiST) to analyze one year of data collected from the Toronto Fire dispatch system (http://www.toronto.ca/fire/cadinfo/livecad.htm).

AGENDA

Pro-Talk [30 min]: Nick Zhu (http://linkd.in/ThZldk), an expert in big data and distributed systems, will present D3 & Crossfilter fundamentals with some visualizations that he has prepared in advance.

Nick is the author of dc.js (http://bit.ly/ThZw8C) (400+ stars on github), a javascript charting library with native Crossfilter support. It's being used by giants such as GE Research and Google.

Live Coding Tutorial [90 min]: As a group, we will modify Nick's sample project to answer the following questions:

1. Which month had the highest volume of crime related incidents?

2. How many trucks responded on average to each type of incident?

3. What was the average number of trucks and duration of incident by Alarm Level? (i.e. Alarm Level 3 fires require 8 trucks on average and last over 5 hours)

REQUIREMENTS

1. Laptop with a web server serving static files from a known location - this could be Apache, Nginx or even something as simple as this (http://bit.ly/VS7svg).

2. A local copy of the dataset (http://bit.ly/YlzOSb).

Windows users: If you're not able to get a web server running locally, make sure you have the latest Firefox installed.

WHO SHOULD ATTEND

This session is intended for people who are already comfortable analyzing and visualizing data and interested in adding a new technology to their toolkit.

ABOUT THE DATASET

The following posts should give you an idea of what's in the dataset:

1: Introduction to the Toronto Fire Dispatch Data (http://bit.ly/URp75h)

2: Capturing and Storing the Data - DIY (http://bit.ly/11PwsYn)