Summer, Sun, and Data Visualization
Chris Main (BS CS, 2013) describes his work for the summer of 2011:
As Bucknell continues to install and test alternative energy sources on campus, the staff at the environmental center realized a need to collect and analyze data from these sources. The major goal of my summer work was to develop an interactive, extensible, web-based system for graphically presenting the data being collected from the current alternative energy sources on campus. These data sources include both semi-real-time and archived data from three photovoltaic arrays and one solar thermal hot water heater. The website I developed allows users to view both real-time (within the previous minute) and archived (from 2007 onward) data in a variety of interactive graphs and downloadable spreadsheets. Accessibility, ease of use, and an organized presentation of data were all major considerations in the design and development of the website. You can see the completed work at: http://www.linux.bucknell.edu/~altenergy.
A number of different technical challenges had to be overcome in the development of this web-based hub for alternative energy at Bucknell. The majority of these challenges arose in the organization of the data that had already been collected. To graph easily multiple sets of data on the same plot requires an organized data structure, one that was not already present in the database that had been collecting data since 2007. A complicating factor was that many data points were missing in the database for dates and times where no data was collected. For this reason, various techniques had to be used to organize and merge the data sets on the fly for generating graphs. Since we would be offering line graphs for monthly data sets, organizing thousands of data points with an acceptable response time was a requirement. When generating spreadsheets, execution time was even more critical, as users could potentially request multiple years worth of data at once, amounting to hundreds of thousands of data points. This resulted in two separate algorithms for organizing data. The graphing algorithm, while slower, could plot points that were taken a few seconds apart with a common time value. The spreadsheet algorithm did not merge data points, but was developed to be much faster so that it could handle large, raw data sets.
This project was developed under the guidance of both the College Engineer, Wade Hutchison, and the Director of the Campus Greening Initiative, Dina El-Mogazi.