Functional programming in Python

Functional Programming in Python
(PYTHON)( )

TODAY'S SLIDES
http://goo.gl/iRXD0Y

(PYTHON)( )
ABOUT ME
+ Colin Su a.k.a LittleQ
+ http://about.me/littleq
+ Productional Experience:
- Erlang
- Python
- Java{,Script}

(PYTHON)( )
PROGRAMMING IN PYTHON
+ Imperative Programming
- Shell Scripting
+ Procedural Programming
- Languages with functions
+ Declarative Programming
- Functional Programming

(PYTHON)( )
FUNCTIONAL PROGRAMMING
+ No state
+ Immutable data
+ Function as ﬁrst-class citizen
+ Higher-order function
+ Purity
+ Recursion, tail recursion
+ ...the list is too long to name here

(PYTHON)( )
IP VS FP
$ ./program1
$ ./program2 --arg=1
$ ./program3
$ ./program1 | ./program2 --arg=1 | ./program3
Imperative Functional
Your Terminal

(PYTHON)( )
FUNCTION AS FIRST CLASS OBJ
def add(a, b):
return a + b
add2 = add
add2(1,2) # 3
def giveMeAdd():
def add(a, b):
return a + b
return add

(PYTHON)( )
LAMBDA SUPPORT
add = lambda a, b : a + b
minus = lambda a, b : a - b
multiply = lambda a, b : a * b
divide = lambda a, b : a / b

(PYTHON)( )
(POOR) LAMBDA SUPPORT
add = lambda a, b:
c = a + b
return c

(PYTHON)( )
(POOR) LAMBDA SUPPORT
add = lambda a, b:
c = a + b
return c
File "test.py", line 29
add = lambda a, b:
^
SyntaxError: invalid syntax

(PYTHON)( )
HIGHER-ORDER FUNCTION
+ Function eats functions, and returns functions
def food():
return "food got eaten"
drink = lambda: "drink got drunk"
def me(*args):
def eat():
return map(lambda x: x(), args)
return eat
act = me(food, drink)
print act()
>>> ['food got eaten', 'drink got drunk']

(PYTHON)( )
MAP/REDUCE/FILTER
+ A war between readability and conciseness
+ To make your code pluggable

(PYTHON)( )
CALCULATE "1++2+3+++4+5"

(PYTHON)( )
IMPERATIVE.PY
INPUT = "1+2++3+++4++5+6+7++8+9++10"
result = 0
for num in INPUT.split('+'):
if num:
result += int(num)
print result
Iterate on the same variable to perform task

(PYTHON)( )
FUNCTIONAL.PY
from operator import add
INPUT = "1+2++3+++4++5+6+7++8+9++10"
print reduce(add, map(int, filter(bool, INPUT.split('+'))))

(PYTHON)( )
FUNCTIONAL PYTHON
MADE SIMPLE

(PYTHON)( )
BUT NOT ALL LANGUAGES
DO...

(PYTHON)( )
IMPERATIVE.JAVA
Multiset<Integer> lengths = HashMultiset.create();
for (String string : strings) {
if (CharMatcher.JAVA_UPPER_CASE.matchesAllOf(string)) {
lengths.add(string.length());
}
}
Looks making sense

(PYTHON)( )
FUNCTIONAL.JAVA
Function<String, Integer> lengthFunction = new Function<String, Integer>() {
public Integer apply(String string) {
return string.length();
}
};
Predicate<String> allCaps = new Predicate<String>() {
public boolean apply(String string) {
return CharMatcher.JAVA_UPPER_CASE.matchesAllOf(string);
}
};
Multiset<Integer> lengths = HashMultiset.create(
Iterables.transform(Iterables.filter(strings, allCaps), lengthFunction));

(PYTHON)( )
Welcome To Functional

(PYTHON)( )
But my typing speed really got increased since using Java.

PARTIAL FUNCTION
APPLICATION
(((a × b) → c) × a) → (b → c)

(PYTHON)( )
PARTIAL FUNCTION
APPLICATION
from functools import partial
def f(x, y, z):
return x + y + z
f2 = partial(f, 100, 10)
print f2(5) #115

(PYTHON)( )
PARTIAL FUNCTION
APPLICATION
+ Keyword-based partial application
from datetime import datetime
def log(message, prefix="", postfix=""):
print prefix, message, postfix
error = partial(log, postfix=datetime.now(), prefix="[ERROR]")
error("something goes wrong")
# [ERROR] something goes wrong 2014-04-10 01:37:07.250509

CURRYING
((a × b × c) → d) → (((a → b) → c) → d)

(PYTHON)( )
CURRYING
+ Simple sum
def simple_sum(a, b):
return sum(range(a, b+1))
>>> simple_sum(1, 10)
55

(PYTHON)( )
CURRYING
+ Squared sum
def square_sum(a, b):
return sum(map(lambda x: x**2, range(a,b+1)))
>>> square_sum(1,10)
385

(PYTHON)( )
CURRYING
+ Square root sum
def sqrt_sum(a, b):
return sum(map(math.sqrt, range(a,b+1)))
>>> sqrt_sum(1,10)
22.4682781862041

(PYTHON)( )
CURRYING
+ Curried sum()
import math
def fsum(f):
def apply(a, b):
return sum(map(f, range(a,b+1)))
return apply
simple_sum = fsum(int)
square_sum = fsum(lambda x: x**2)
sqrt_sum = fsum(math.sqrt)
print simple_sum(1,10) # 55
print square_sum(1,10) # 385
print sqrt_sum(1,10) # 22.4682781862

(PYTHON)( )
CURRYING
+ Combined with partial
def fsum(f):
def apply(a, b):
return apply
from operator import mul
mul_sum = fsum(partial(mul, 2))
print mul_sum(1,10) # 110

(PYTHON)( )
DECORATOR
+ Another way to curry
import math
def fsum(f):
def apply(a, b):
return apply
@fsum
def sqrt_sum(x):
return math.sqrt(x)
print sqrt_sum(1,10) # 22.4682781862

(PYTHON)( )
GENERIC CURRY IN PYTHON
import functools
def curry(func):
def curried(*args, **kwargs):
if not args and not kwargs:
return func()
return curry(functools.partial(func, *args, **kwargs))
return curried
@curry
def add(a, b, c, d, e, f, g):
return a + b + c + d + e + f + g
add12 = add(1)(2)
add1234 = add12(3)(4)
add1234567 = add1234(5)(6)(7)
print add1234567() # 28

(PYTHON)( )
TAIL RECURSION
+ No grammar-level support, just simulation
def printAll(strings):
if strings:
print strings[0]
return printAll(strings[1:])
printAll([1,2,3,4,5])
Python 2
def printAll(strings):
if strings:
head, *tail = strings
print(head)
return printAll(tail)
printAll([1,2,3,4,5])
Python 3

(PYTHON)( )
FUNCTIONAL PYTHON?
Pros
+ First-class function
+ lambda
+ built-in map/ﬁlter/reduce
+ functools
+ generators as lazy-evaluation
Cons
+ non-pure
+ lambda (?)
+ memory-cost operations
+ No optimization for tail recursion
+ No pattern matching

(PYTHON)( )
CONCLUSION
+ Python is not for replacing any functional
language
+ But a good way to begin functional
programming

(PYTHON)( )
ALL CODE SAMPLES ON GIST

Functional programming in Python

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Functional programming in Python (20)

More from Colin Su (20)

Recently uploaded (20)

Functional programming in Python