NBA Data Analysis

A simple project for relative beginners at Scala working on data engineering tasks.

The task is to create a service that analyses play-by-play data to create a summary of the game that is known as a box score. The NBA has full data from this game, including box score, play-by-play data, and highlights, at https://watch.nba.com/game/20200306/MILLAL Our box score is a bit simpler than the real box score, but you can look at the real box score to get a feel for what we're after.

The play-by-play data is in a CSV file that we have included in the repository. The data was obtained from the sample log at https://www.bigdataball.com/nba-historical-playbyplay-dataset/.

Components

• EventService - The main application you are to build up. Its job is to:
  1. Read events until there are no more.
  2. Collate a BoxScore from the events.
  3. Post the BoxScore to the BoxScoreService, ensuring there are no errors.
• DataSource - Translates raw CSV data to a structured event format.
• BoxScoreService - Contains the definition of a BoxScore and a (fake) method to post it to a remote endpoint.

Tasks

1. Data Modeling

   1. Examine the data under src/main/resources to get a sense of the structure and fields.

   2. Decide how to encode the CSV rows as a data structure (algebraic data type). Create this type as DataSource.Event.

   3. What methods are available to use from the CsvReader? How do they work? You can use the console, via sbt console, etc. The console (REPL) has tab-completion, which is very handy:

      > sbt console
      ...
      
      scala> import io.innerproduct.nba._
      import io.innerproduct.nba._
      
      scala> val reader = DataSource.reader
      reader: kantan.csv.CsvReader[kantan.csv.ReadResult[List[String]]] = kantan.codecs.resource.ResourceIterator$$anon$3@75331de9
      
      scala> reader.<TAB>
      ... lots of methods here ...        
      
      scala> reader.next
      res2: kantan.csv.ReadResult[List[String]] = Right(List(GAME-ID, Season that this data sets belongs to., Game
      ...
      

   4. Implement DataSource.event() to translate an event from a row of the CSV.

2. Reading events.
   1. In EventService, read all the available events. Decide how to handle errors during reading, if any.
   2. Design the transformation of DataSource.Event values into the datatypes of BoxScoreService. How do you accumulate the necessary state? How do you validate the necessary inputs and outputs?
   3. Sidebar: Validated.
3. Post valid events.
   1. Invoke BoxScoreService.post and handle successes and errors.