aafa / markit

Historical financial data processing in a pure reactive fashion

Task

Prerequisites Get the csv stock daily data for the past year. Here is the url example https://www.google.com/finance/historical?output=csv&q=GOOG

Task Write Scala functions that will return, feel free to adapt function signatures, these are just indicative

/* 1 - 1 year historic prices given a ticker */
`def dailyPrices(ticker: String) : List[Double]` 
 
/* 2- daily returns, where return = ( Price_Today – Price_Yesterday)/Price_Yesterday */
`def returns(ticker:String) : Seq[Double]`
 
/* 3 – 1 year mean returns */
`def meanReturn(ticker:String): Double`

Objectives

• solve the problem in an elegant and effective FP way
• make fun along the way, use Scala.js, try something new with it

Work process

Initially I thought of doing simple scala.js ui in top of this solution for demo purposes. But it turns out to be a significant pain in the a frustration to force this thing to work. CORS issue is the one in particular.
Due to time constraints I decided to skip this ui part and roll out CLI solution as it suffice to demonstrate working prototype.

fs2 streams has been used to accommodate for possible large data amounts in a reliable and fp-pure fashion.

As the result we have end-to-end streaming pipeline that is well ready to process data of any size.

Test and Run

• sbt test runs all the tests
• sbt "run MSN" executes queries for MSN ticket