Functional Programming

What is it?

It's about values

• collections
• records / types
• scalars (int, float, char, etc.)
• functions!

Values lead to expressions

• double(x) = x * 2
• double(4) = 4 * 2
• double(4) = 8
• 8

Functions as values lead to higher-order functions

• multiply-by(n) = \x -> x * n
• double = multiply-by(2)
• map(double, [1,2,3]) = [double(1), double(2), double(3)]
• [\1 -> 1 * 2, \2 -> 2 * 2, \3 -> 3 * 2]
• [1 * 2, 2 * 2, 3 * 2]
• [2, 4, 6]

It's about separating state

• pure functions
• values
• impure functions
• mutation
• IO

What's the big deal?

Loose coupling / High cohesion

Isn't this what it's always about?

Referential Transparency

How much of the system do you need to have in your head to understand a subset of the code?

Generalization / Reuse

It is better to have 100 functions operate on one data structure than 10 functions on 10 data structures.

Alan Perlis

Tips For Learning FP

Practice time

1. Make it work
2. Refactor for readability
3. Refactor for concision
4. Refactor for speed
5. Rewrite a completely different way

Project time

1. Make it work
2. Refactor for readability
3. Refactor when you have a deeper understanding

Learn the core functions

• map
• reduce
• anonymous functions
• collection libraries
• recursion? - surprisingly rare

Separate state and side effects

• Functions that create data structures
• Procedures that do IO
• These two go a long way

Ask for help

• FP programmers are excited that you are interested!
• FP languages have very strong idioms. Learn them.
• FP languages often have performance pitfalls. Ask about them.

Try them out in your environment

• FP principles are not necessarily at odds with other paradigms
• most languages support higher functions
• most languages have higher order collection libraries

Be patient

• FP isn't hard. Learning something new is.