# A4 - Regression. The price of a ticket...solved

A4 - Regression. The price of a ticket depends, in part, on the fuel consumed, figuring out which airline is cheapest requires a little bit of work. Write a MapReduce job that ranks carriers and plots the evolution of prices of the least expensive carrier over time.

Fine Print: (0) Group assignment, two students. (1) Take as input "-time=N" where N is a natural number representing scheduled flight minutes. (2) For each year and each carrier, estimate the intercept and the slope of a simple linear regression using the scheduled flight time to explain average ticket prices. For a given year, compare carriers at N by computing intercept+slope*N. The least expensive carrier is the carrier with lowest price at N for the most years. (3) Plot, for each week of the entire dataset, the median price of the least expensive carrier. (4) The reference solution is 300 LOC of Java with two MR jobs pipelined. (5) Use the full airline dataset form s3://mrclassvitek/data -- it contains 337 files! (6) Produce a report that discusses your results and includes graphs for N=1 and N=200. (7) The reference solution has 300 lines of Java.

Fine Print: (0) Group assignment, two students. (1) Take as input "-time=N" where N is a natural number representing scheduled flight minutes. (2) For each year and each carrier, estimate the intercept and the slope of a simple linear regression using the scheduled flight time to explain average ticket prices. For a given year, compare carriers at N by computing intercept+slope*N. The least expensive carrier is the carrier with lowest price at N for the most years. (3) Plot, for each week of the entire dataset, the median price of the least expensive carrier. (4) The reference solution is 300 LOC of Java with two MR jobs pipelined. (5) Use the full airline dataset form s3://mrclassvitek/data -- it contains 337 files! (6) Produce a report that discusses your results and includes graphs for N=1 and N=200. (7) The reference solution has 300 lines of Java.

You'll get a 14.5KB .ZIP file.