Here is a quick overview of HW3:

1. Solve the mtcars exercise:

   a. Split mtcars into 3 files. (Use the code I provided in
      examples/5mtcarsPractice5mtcarsPractice/instructions.txt.)
   b. To process one file, use a pipeline like the one in
      examples/5mtcarsPractice5mtcarsPractice/example.sh.
   c. To get (b)'s processing to run in parallel across the three
      files, start from examples/4jobArray.

2. Answer questions about air travel from data in 22 files, 1987.csv
   through 2008.csv:

   (No "pre-parallel" step is necessary.)
   a. To process one file like 1987.csv:
     - download the large file (~ 500MB)
     - unzip it
     - write a pipeline to select the required subset of columns
       (DayOfWeek, DepDelay, Origin, Dest, Distance) and rows
       (Origin=MSN) and write a much smaller file (~ 0.1 MB) for later
       processing.
   b. Same idea as 1(c).
   c. Then combine the 22 smaller files into a single file (~2.5 MB).
      Process it twice:
      - Use a bash pipeline to answer "How far can you get from MSN
        in one flight?"
      - Use R to answer "What is the average departure delay for each
        day of the week?"
   d. Delete the large files (like 1987.csv) to save space.

   I recommend starting from the same example.sh and 4jobArray
   examples as in (1).