By now, you should be more comfortable using Spark's Dataset API. Our little program is focused on fetching BCT/USD transactions, but it could be interesting to enhance it. For instance, you could fetch and save other currency pairs, such as ETH/EUR or XRP/USD. Use a different cryptocurrency exchange. This would allow you to compare prices in different exchanges, and possibly work out an arbitrage strategy. Arbitrage is a simultaneous purchase and sale of an asset in different marketplaces to profit from an imbalance in the price. You could get data for traditional currency pairs, such as EUR/USD, or use Frameless to refactor the Dataset manipulations and make them more type-safe. See the website for further clarification https://github.com/typelevel/frameless.
In the next chapter, we are going to exploit saved transaction data to perform some analytics...
