Library Loans by Media Modernity by Darren Hardy

This visualization categorizes media into 4 classes based on the modernity of the media (e.g., Books are in the first class, and DVDs are in the last class). It draws a histogram of the checkout volume per category per day for the entire corpus from June 2005 to January 2009.

Written for MAT 259 (Legrady): Linear Data Flow Project, 22 January 2009.

Sketch

Screenshots

Applet

Run the applet (pops new window).

Code

Data processing

• Use wget to mirror tango XML data between June 2005 and January 2009. Output is 40GB.
• Wrote shell script to convert XML from UTF-16 to UTF-8. Output is 20GB.
• Wrote perl script to convert XML into UTF-8 flat CSV file. Output is 1.3GB.
• Wrote perl script to summarize total transactions per itemtype per day. Output is 600MB, summarized.txt.gz.
• Wrote Processing script to calculate histograms and media count table. Output is 0.1MB, histogram.txt, mediacount.txt.
  Source: HardyProjectOneData.pde, ItemType.pde
  Data: mediamap.txt

Data visualization

setup()

• Draws title, axis labels, and informational text on screen
• Loads histogram data into a SortedMap with date as key and histogram as value

draw() for each frame

• Fetches the earliest N histograms by popping the first date key from the SortedMap
• Draws N semi-transparent histograms per frame
• Draws progress bar as semi-transparent

Source: HardyProjectOne.pde
Data: histogram.txt, mediacount.txt