path: root/advanced_python/16_generators.ipyml

cells:

- markdown: |
    # Advanced Python: Iterators and Generators

    ### Prabhu Ramachandran
    ### The FOSSEE Python group &
    ### Department of Aerospace Engineering
    ### IIT Bombay

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    ## Background

    - Container objects present a similar interface
    - They can all be looped over using `for`

  metadata:
    slideshow:
      slide_type: slide

- code: |
    for i in [1, 2, 3]:
        print(i)

    for key in {'a': 1, 'b': 2}:
        print(key)

    for char in 'hello world':
        print(char)

    for line in open('file.txt'):
        print(line, end='')

- markdown: |
    ## Background

    - Similar interface for different objects
    - These objects are all **iterable**
    - They are all **iterators**

  metadata:
    slideshow:
      slide_type: subslide

- markdown: |
    ## Introduction

    - This is a powerful feature/abstraction
    - Easy to create your own iterators
    <br/>
    <br/>

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    - A **generator** makes it easy to create an iterator
    - How would you write a `range` function?
    <br/>
    <br/>

    - Use the `yield` keyword

  metadata:
    slideshow:
      slide_type: fragment

- code: |
    def counter(n):
        count = 0
        while count < n:
            yield count
            count += 1

  metadata:
    slideshow:
      slide_type: slide

- code: |
    c = counter(2)

- code: |
    next(c)

- code: |
    next(c)

- code: |
    next(c)

- code: |
    for i in counter(5):
        print(i)

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    ## Observations

    - Note that `yield` resumes the function
    - The function/method remembers its state
    - It can yield any object
    - An empty `yield` is also acceptable
    - You can have multiple yields

  metadata:
    slideshow:
      slide_type: slide

- code: |
    def infinite_generator():
        while True:
            yield

  metadata:
    slideshow:
      slide_type: fragment

- markdown: |
    ## Exercise: simple generator

    Write a simple generator that generates the first `n` odd numbers. Call this
    function `odd(n)`.

  metadata:
    slideshow:
      slide_type: slide

- code: |
    for x in odd(5):
        print(x)

- markdown: |
    ## Solution

  metadata:
    slideshow:
      slide_type: slide

- code: |
    def odd(n):
        for i in range(n):
            yield 2*i + 1

- markdown: |
    ## Exercise: Fibonacci generator

    Write a simple generator that returns the first `n` numbers of the Fibonacci
    sequence, call this function `fib(n)`.

  metadata:
    slideshow:
      slide_type: slide

- code: |
    for x in fib(10):
        print(x)

- markdown: |
    ## Solution

  metadata:
    slideshow:
      slide_type: slide

- code: |
    def fib(n):
        a, b = 0, 1
        yield 0
        for i in range(n-1):
            yield b
            a, b = b, a + b

- markdown: |
    ## Creating iterable objects

    - What if you want to create an object that is iterable?
    - Overload the `__iter__` and/or `__next__` methods.
    - Recall our `Zoo` class?

  metadata:
    slideshow:
      slide_type: slide

- code: |
    class Zoo:
        def __init__(self, *animals):
            self.animals = list(animals)

- markdown: |
    We want the following to work,

  metadata:
    slideshow:
      slide_type: slide

- code: |
    c = Animal('crow')
    m = Mammal('dog')
    z = Zoo(c, m)

    for animal in z:
        animal.greet()

- markdown: |
    ## Making `Zoo` iterable

    - Overload the `__iter__` to return an iterable
    - An iterable should have a `__next__` method

  metadata:
    slideshow:
      slide_type: slide

- code: |
    class Zoo:
        def __init__(self, *animals):
            self.animals = list(animals)

        def __iter__(self):
            return iter(self.animals)

- code: |
    class Animal(object):
        def __init__(self, name):
            self.name = name
        def greet(self):
            print(self.name, "says hi!")

    class Mammal(Animal):
        pass

  metadata:
    slideshow:
      slide_type: slide

- code: |
    c = Animal('crow')
    m = Mammal('dog')
    z = Zoo(c, m)

    for animal in z:
        animal.greet()

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    ## Observations

    - Just returns `iter(self.animals)`
    - `self.animals` is a `list` and calling `iter` makes it an iterator

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    ## A more complex example

    - What if we cannot just return an internal list?
    - We must provide a `__next__` method
    - To stop the iteration: `raise StopIteration`

  metadata:
    slideshow:
      slide_type: slide

- code: |
    class SimpleRange:
        def __init__(self, n):
            self.n = n
            self.count = 0
        def __iter__(self):
            return self
        def __next__(self):
            if self.count == self.n:
                raise StopIteration
            self.count += 1
            return self.count

  metadata:
    slideshow:
      slide_type: subslide

- code: |
    r = SimpleRange(5)
    for i in r:
        print(i)

  metadata:
    slideshow:
      slide_type: subslide

- markdown: |
    ## Observations

    - Useful when you want an object to be iterable
    - Notice the `__next__` method
    - Notice the `raise StopIteration`
    - Easier to use functions/methods with `yield`

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    ## Summary

    - Easy to use and write
    - Allows one to abstract a loop
    - `yield` for functions and methods
    - Overload `__iter__, __next__` for objects

  metadata:
    slideshow:
      slide_type: slide

- markdown: |
    ## Exercise: Fibonacci object

    Create a class that when instantiated with an argument `n` can be used as a
    sequence of the first `n` numbers of the Fibonacci sequence, call this
    class `Fib`.

  metadata:
    slideshow:
      slide_type: slide

- code: |
    f = Fib(5)
    for x in f:
        print(x)

- markdown: |
    ## Solution

  metadata:
    slideshow:
      slide_type: slide

- code: |
    class Fib:
        def __init__(self, n):
            self.count = n
            self.a, self.b = 0, 1
        def __iter__(self):
            return self
        def __next__(self):
            if self.count == 0:
                raise StopIteration
            old = self.a
            self.count -= 1
            self.a, self.b = self.b, self.a + self.b
            return old

- code: |
    f = Fib(10)
    for i in f:
        print(i)

  metadata:
    slideshow:
      slide_type: slide

# The lines below here may be deleted if you do not need them.
# ---------------------------------------------------------------------------
metadata:
  celltoolbar: Slideshow
  kernelspec:
    display_name: Python 3
    language: python
    name: python3
  language_info:
    codemirror_mode:
      name: ipython
      version: 3
    file_extension: .py
    mimetype: text/x-python
    name: python
    nbconvert_exporter: python
    pygments_lexer: ipython3
    version: 3.5.2
  rise:
    scroll: true
    transition: none
nbformat: 4
nbformat_minor: 2