Visualization of Household Utility Data



Introduction
This visualization was built for the first assignment of CS 424: Visualization and Visual Analytics, taught by Professor Andrew Johnson in the Computer Science Department at UIC.  The goal of the project was to take 11 years of Professor Johnson's daily household utility data, collected between 1999-2012, along with monthly average temperatures and notable household events - remodeling, vacations, etc - and display this information in an interactive and accessible format. The project was completed using Processing.  

Downloading and running the application
Standalone applications:  To run, double click on the .exe or .app file. 
Source code and application for mac, windows and linux 
    Here
ControlP5 library needed to run the source- download, unzip and place in the libraries directory in the Processing sketchbook 
    Here


How to use the application

The main body of the application contains three plots showing average daily consumption of water, electricity and natural gas in each month of the entire 11-year time period.  When you hover over a point in the plot, you can compare that month's usage to, for instance, the average usage in September and the average usage in all the months.  This allows the user to see whether consumption in this particular month is unusually high or low, taking into account seasonal variation.  





 
Initially, all months are displayed.  The user can select a time period, using the time selection bar at the bottom, and then press the 'Zoom' button to see just that period.  For instance, if the user is interested in consumption between Jan 2003 and Jan 2005, they can select this period by dragging the ends of the slider bar and then can hit 'zoom' to view this period in detail.  A reset button will then appear, allowing the user to return to the initial view of the entire time period. 




To view the consumption data in relation to temperature, the user can click on the 'Draw Temp Curve' checkbox, and the average monthly temperature will be plotted atop the consumption data.  Given the relationship between weather conditions and utility consumption, this will put the usage data in a useful context. 




The relationship between monthly average temperature and consumption can be further explored by clicking on the 'Show Vs Time' button.  The initial plot will fade, revealing a scatterplot showing consumption vs temperature.  Hovering over any point in the scatterplot will show month associated with that value. This view shows outliers fairly clearly - eg. hot months with low electricity usage, for instance. 




Further context on the data can be found by relating the data points to significant household events- such as household remodeling projects or vacations- where utility consumption would be expected to change.  To view this events, click on the 'Highlight Household Events' bar and this time period will be illuminated in the data set.  


Some unusual consumption patterns can be explained by these household events, such as water consumption increasing during the time period in which Prof. Johnson enlarged his pond. 


Further information about this event is provided in a small plot in the lower right hand corner comparing the consumption of relevant utilities before, during and after the event.  This way, it is evident whether this event had a lasting or temporary impact. For example, it is clear from the plot of water consumption below that enlarging the pond had a significant effect on water consumption during the remodeling period, and is associated with a long term increase in water usage. 

Project completed September 2010.





Comments