By: Nirari Barm, CS Graduate Student, University of Illinois at Chicago

Screenshot of the final version of my graphing tool

Visualization Tool Features

I designed this interactive visualization tool to graph data over time.  In this project, I’m using it to graph my professor’s household utility data over almost 12 years (links to the application are at the end of this post).  Before you visit my data analysis page, here are the main features of my tool:

There are three household utility data plots types to choose from this tap menu:

Data Tap Menu

With these toggle buttons you can show/hide visual data features:

Data Features Toggle Buttons

Temp: plots the temperature bar graph behind the data

Month: plots the average monthly usage line

Year: plots the average yearly usage curve

Season: plots the average seasonal usage graph

Events: shows dates of changes in usage and home improvements marked with triangles on the plot and dots on the slider bar

MouseX: draws a vertical line on the plot at x-position of mouse pointer

MouseY: draws a horizontal line on the plot at y-position of mouse pointer

Area: fills the area under the data curve

Curve: connects the data plot points with a curve

Points: shows the data plot points

Δx/Δ°T: plots estimated changes in data for heating/cooling per change from normal temperatures

Approx: plots predicted usage data for the next months (Aug, Sept, Nov, and Dec of 2010)

To view more info about a data point, simply highlight it with the pointer:
Data Highlight

Data Highlight

You can also highlight events indicated by yellow triangles:
Event Highlight

Event Highlight

I placed an info box on the right of the plot to give an estimated utility usage of regular activities in a home:
General Info Box

General Info Box

The slider bar under the plot can be resized and moved for better view of the data:
Slider Bar

Slider Bar

The white head button sets the beginning date and the gray end button for the ending date for data points plotted.

The yellow dots represent the event dates for when changes to the house or its activities occurred.

The green bar highlights the time for the predicted data values of future months.

I added a cluster control bar to average data in groups for better comparison between seasons’ and years’ usage:
Cluster Bar

Cluster Bar

Heres what it looks like with a cluster of 12 months (1 year):

Clustering by One Year

Clustering by One Year

The areas under each year of data are easily compared this way and in other cluster sizes.

The highlights above sum up the main features of my interactive visualization tool.  Feel free to run it on your browser or computer and post comments or questions if you like.  You can also visit my data analysis page for my report on the project using my tool.

Click Here! to download the application (Win, MacOSX, and Linux).
Download Mirror 1
Download Mirror 2
Click Here! to run the application on your browser (Java Required).
App Mirror 1

© 2010 Nirari Barm