Downloaded 17 times
05 functional programming
- 1.
- 2. What’s wrong withimperative code?
- 3. What’s wrong withimperative code?
- 4. What’s wrong withimperative code?
- 5. What’s wrong withimperative code? • It is repetitive • It is lengthy • (higher chances to introduce bugs) • (harder to read) • It describes what to do and how • (mixing low-level and high-level concepts)
- 6. Let’s try tosolve these problems
- 7.
- 8.
- 9.
- 10. Delegates are immutable Youcan only change the delegate variable
- 11. Delegate knows itstarget
- 12. Look, we manipulatecode as if it was data! Nothing new though (we had interfaces for ages)
- 13. So we can • Store it in variable • Store it in a field or property • Pass it as a method parameter • Return it as a method result
- 14. Functions (delegates) are firstclass citizens Both pure and impure
- 15. Pure vs side-effects •Pure functions: – Maht.Sin(X) – String.Length – List<T>.IndexOf(T) • Functions with side effects: – List<T>.Add(T) – DateTime.UtcNow – Console.WriteLine()
- 16.
- 17. Consider the problem •Initiate processing • When some condition occurs, we want to execute some code
- 18.
- 19.
- 20.
- 21. Nothing really complicatedso far Let’s go deeper
- 23. Separate method --not needed
- 24. Lambda expressions (lambdafunctions)
- 25. Return value fromdelegate
- 26.
- 27.
- 28. How to makeit more flexible?
- 29. This is calledclosure (lexical closure) The function captures variables from the scope
- 30.
- 31.
- 32. Function can evencapture itself
- 33.
- 34.
- 35. We’ve looked atexamples of higher-order functions They take one or more functions as an input or return a function
- 36. Be comfortable withcreating and passing lambdas Understand closures This is very powerful and widely used idiom
- 37.
- 38.
- 39.
- 40.
- 41.
- 42. Currying Transform f: (X * Y) -> Z Haskell Curry into f: X -> (Y -> Z) Input: Function with N parameters Output: chain of N functions with 1 parameter each
- 43.
- 44. How to computeFibonacci(N) What’s wrong with this? It is exponential (slow) link
- 45.
- 46.
- 47. Main ideas • Functionsas first class citizens • Compose complex functions from simple • Create specific functions out of generic ones • More generic and reusable code with higher order functions • For less code and less bugs
- 48.