AnalyticsDojo

Introduction to Python - Functions

introml.analyticsdojo.com

7. Functions#

  • Why functions?

  • (Need to know) Indentation in Python

  • Predefined functions

  • Custom functions

  • Lambda functions

  • Exercises

7.1. Why Functions?#

  • Code reuse.

  • Abstract away complexity.

  • Simple, efficient robust code.

  • Specific functional programming languages like Lisp & Haskell built around functional programming, which enforces great practices.

  • Read more about functional programming in Python here.

7.2. Predefined Functions#

  • We have used predefined functions in earlier exercises

  • Python has predefined functions for embedded data structures like variables and lists.

  • Packages like Numpy and Pandas include functions for working with those data

  • We can string functions together ex. b = np.arange(15).reshape(3, 5)

#Simple Rounding Functions
a=3.14
a=round(a)
print(a)

#We can string functions together
import numpy as np
b = np.arange(15).reshape(3, 5)
print(b)

7.3. Custom Functions in Python#

  • The statement def introduces a function definition, which is followed by the function name and the list of pramaters that are passed to the function as well as a colon to end the line.

  • The second line with three quotes (“””) is called a docstring which can be used to automatically generate documentation.

  • Values used in intermediate calculations within a function are local variables to the function and not global variables.

  • The return statement returns a value to a passed value.

  • Execution of the function requires passing parameters. For example, squared(5) as shown below.

testvalue=0
# Defining New Functions

def squared(x):
    """Function to square a value if passed another value."""  #This is a docstring
    testvalue=10 #Note this is not stored outside of the function but is a local variable which is not returned.
    x=x**2
    return x 
print (squared(5), testvalue)

7.4. Functional Programming#

  • Functions are, like everything else in Python, object.

  • Functions can passed around just like any other value.

  • That means we can do really cool things, like pass a function to a function.

print(squared)

y = squared
print (y)

print (y(5))

7.5. Applying a Function to Each Element of a Collection with Map#

  • We can apply a function to each element of a collection using the built-in function map().

  • Provided using the form map(Function, Sequence).

  • This will work with any collection: dict, list, set, and tuple. (Tuples are immutable data structures similar to lists that we won’t really be using).

exampleList=[1, 2, 3, 4]
list(map(squared, exampleList)) #While Python 2 returned a list, Python 3 will return a map object.

7.6. Applying a Function to Each Element of a Collection with List Comprehensions#

  • Because this is such a common operation, Python has a special syntax to do the same thing, called a list comprehension.

  • Can change whether the result is a set or list by changing brackets.

  • If we want a set instead of a list we can use a set comprehension or dictionary comprehension

#List Comprehensions
list1=[squared(i) for i in [1, 2, 3, 4]]
print(list1)

{squared(i) for i in [1, 2, 3, 4]}

7.7. Applying a Function to Subset of a Collection with List Comprehensions#

  • List Comprehensions can be nested

  • The if statemen can be added to make the function apply to only some.

lista= [(x, y) for x in ['a','b','c'] for y in ['c','d','e'] if x != y]
print(lista)

#This is a comparable for loop. 
listb=[]
for x in ['a','b','c']:
    for y in ['c','d','e']:
        if x != y:
            listb.append((x, y))
        
print(listb)

7.8. Cartesian product using List Comprehensions#

Cartesian Product By Quartl (Own work) [GFDL (http://www.gnu.org/copyleft/fdl.html) or CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0)], via Wikimedia Commons


- The [Cartesian product](https://en.wikipedia.org/wiki/Cartesian_product) of two collections $X = A \times B$ can be expressed by using multiple `for` statements in a comprehension. -This can be thoguht of as nested for loop. 
A = {'x', 'y', 'z'}
B = {1, 2, 3}
{(a,b) for a in A for b in B}

7.9. Cartesian products with other collections#

  • The syntax for Cartesian products can be used with any collection type.

first_names = ('Steve', 'John', 'Peter')
surnames = ('Smith', 'Doe')

[(first_name, surname) for first_name in first_names for surname in surnames]

7.10. Anonymous Functions: lambda expressions#

  • While we can create named functions with def, we can also write anonymous functions that do not necessarily have a name.

  • They are called lambda expressions (after the \(\lambda-\)calculus).

  • Useful if creating a function that is only going to be used once.

  • Previously we created a squared function. This creates an anonymous lambda function.

y=map(lambda x: x ** 2, [1, 2, 3, 4])
print(list(y))

7.11. lambda : Filtering LIst#

  • We can filter a list by applying a predicate to each element of the list.

  • A predicate is a function which takes a single argument, and returns a boolean value.

  • We will be using a similar procedure to filter DataFrames.

lista=[-5, 2, 3, -10, 0, 1]
listb=list(filter(lambda x: x > 0, lista))
print(listb)

#Here we could rereate an equavalent filter. 

def postive_numbers(listz):
    return [i for i in listz if i>0]  #Here we are using list comprehension.
postive_numbers(lista)

7.12. Example Filter String#

We can use both filter() and map() on other collections such as strings or sets.

list(filter(lambda x: x != ' ', 'hello world'))

list(filter(lambda x: x > 0, {-5, 2, 3, -10, 0, 1}))

7.13. Filtering using a List Comprehension#

  • Again, because this is such a common operation, we can use simpler syntax to say the same thing.

  • We can express a filter using a list-comprehension by using the keyword if:

data = [-5, 2, 3, -10, 0, 1]
[x for x in data if x > 0]

#We can also filter and then map in the same expression:
from numpy import sqrt
[sqrt(x) for x in data if x > 0]

7.14. Big Data#

  • The map() and reduce() functions form the basis of the map-reduce programming model.

  • Map-reduce is the basis of modern highly-distributed large-scale computing frameworks.

  • It is used in BigTable, Hadoop and Apache Spark.

7.14.1. Modules#

  • Functions are great, but you might not want to repeat same function in different programs.

  • To faciliate code reuse, Python modules can be imported.

  • Must have modeules placed somewhere they are in the PYTHONPATH (a list of directory names, with the same syntax as the shell variable PATH).

  • Review the fibo.py file, which contains a function for a Fibonacci series

# A Fibonacci series is a series of numbers in which each number ( Fibonacci number ) is the sum of the two preceding numbers.
import fibo
fibo.fib(1000)

7.14.2. Working with Modules in Jupyter Notebooks#

  • If a model is loaded, re-running the import command won’t update it.

  • Instead, we need to reload it.

  • Alternately, one could restart the kernal.

def get5():
    return 5

!wget https://raw.githubusercontent.com/rpi-techfundamentals/spring2019-materials/master/03-python/fibo.py

#@title
import importlib
import fibo
importlib.reload(fibo)  # 
fibo.get5()

This work is licensed under the Creative Commons Attribution 4.0 International license agreement. Adopted from materials Copyright Steve Phelps 2014