This repository contains code developed during the Scala Functional Programming capstone project https://www.coursera.org/learn/scala-capstone/home/welcome
I chose to use Spark to scale the data processing as it is a good fit for massively parallel tile generation. As I didn't have access to a Hadoop cluster, I ran it in stand-alone mode on a few nodes of my work cluster. The resulting processing time was around 20-30mins.
If you look back through the commits you will get a good idea of how I targetted each milestone in turn before refactoring into a more coherent overall structure.
The resulting web application is available on the web at: