Categorize transcript data by course

In raw-data/transcript, collect unique course titles and recode them to "math", "science", etc as appropriate. Ultimately, it may be more useful to categorize these by academic vs. non-academic (e.g. exclude physical education), and core (math, english, science, social studies) vs. non-core subjects (all other).

To easily summarize the fields from the command line, can do e.g.:
cut -d "," -f6 sys2012_final.csv | sort | uniq > unique-course-values_sy2012.txt
(where the 6th column has course titles).