Book review: “Haskell Financial Data Modeling and Predictive Analytics”
A few weeks ago, I’ve been asked to review the book “Haskell Financial Data Modeling and Predictive Analytics” by Pavel Ryzhov. I never got such a request before, so I thought “Hey, why not have a look at it”. So I accepted to write a review of the book, once I read it.
Financial Data Modeling?
You might ask what I have to do with financial data modeling. Well, I don’t know either but it’s never too late to explore new fields… ;-)
The Haskell Platform
Haskell is a functional programming language, that has been around for more than 20 years. And as functional programming languages are being hyped everywhere these days, learning another one can’t be wrong.
I already had a look at Haskell in August and September 2012 when I started reading the free introductory book “Learn You a Haskell for Great Good!” by Miran Lipovača, which I really enjoyed. It’s a brilliant introduction to Haskell and its style is really fun to read. But this review is not about Miran’s book…
The Book
After a short chapter on how to setup all necessary tools on various platforms (which tends to focus on Mac OS X), the author introduces the most important principles of Haskell without going too much into detail. If you haven’t seen any Haskell before, the introduction to the language might be a little too short. Probably enough for a financial programmer who wants to get productive.
However the introduction was ideal for me as I haven’t programmed much Haskell since “Learn You a Haskell”. You’ll also be confronted with the first major shortcomings of the book: Almost every code example isn’t part of the book itself, but is only available as part of a ZIP archive that can be downloaded along with the e-book. This might be acceptable when you are reading the book with a normal PC. When you are reading the book with an e-book reader, you can’t access the code examples. Too bad.
During my read I learned the most by studying the code examples, since the explanations often don’t go much into detail. It’s hard to follow the author’s chain of thought even when you have the examples at hand.
The next chapter, “Getting your Hands Dirty”, starts by explaining (file) I/O and parsing in Haskell. It introduces the Attoparsec library which can handle CSV files in a type-safe manner. Additionally, the concept of managing data by using an object relational mapper is introduced and shown by examples.
The following chapter is all about “measuring tick intervals”. It provides solutions for maximum likelihood estimation as well as other models for several financial processes and the secant root finding algorithm. Towards the end of the chapter, QuickCheck is introduced, which is a great library for testing Haskell programs automagically. In QuickCheck, a programmer can define properties that a function must satisfy and QuickCheck randomly generates cases the function will be tested on. This is a great addition to conventional unit testing. I also found ScalaCheck, the Scala version of QuickCheck, very helpful for writing Scala test cases.
I had a hard time reading chapter 4 and 5, since they are diving deep into financial mathematics. That’s why I can hardly give a rating of the quality of these chapters.
One interesting part can be found in chapter 5. Here, parallel computations are introduced and shown by example.
Chapter 6 – the last one – gives an introduction to Cabal, the Haskell build system. It can be used to build a project, manage dependencies, run tests and more. So this is pretty much like Maven/Gradle/SBT for Java and Scala or NPM for Node.js. It’s an interesting way to round up the book with a chapter about Cabal – I would have expected it at the beginning – but it’s important to know Cabal if you want to get started with Haskell.
Conclusion
I would have found it better if the author had explained his code in more detail, but that’s just my opinion. The book is not for programmers who don’t have anything to do with financial programming. Wow, who could think of that?! But it seems to be a neat introduction to financial programming in Haskell, if you are willing to learn the language.
Links
- Pavel Ryzhov: “Haskell Financial Data Modeling and Predictive Analytics”, Packt Publishing, 2013.