Capital One Transaction Challenge (mindsumo.com) (this needs some work, code is too slow, apply numpy vectorization to speed up)
Online businesses of all sorts are increasingly reliant on credit card transactions and the data they create. Information on where clients are, their likely purchasing behaviors, and other bits of data can be used by business to improve their practices. Taking the raw data and turning it into useful information is no easy task though, which is why many engineers find themselves processing and manipulating it on a regular basis.
###Write a script: Using the transactions data attached below, write a script in Java, Python, C/C++, or JavaScript that outputs a list of subscription IDs, their subscription type (daily, monthly, yearly, one-off), and the duration of their subscription.
###Bonus Questions (not required):
-
Give annual revenue numbers for all years between 1966 and 2014. Which years had the highest revenue growth, and highest revenue loss?
-
Predict annual revenue for year 2015 (based on historical retention and new subscribers)