python crash course - Learning from Echoes

a very short introduction to python

variables and data types number, boolean, string, list, dictionary
conditionals and loops if, for, while
functions
classes
standard library

numpy/scipy numerical/scientific computing
pandas data analysis and manipulation
matplotlib visualization
statsmodels statistical modeling
sklearn machine learning
keras/tensorflow pytorch deep learning

print("Hello world!")

Hello world!

Basics¶

variables and data types¶

numbers¶

in python data types are implicitly declared: depending on how you write a number it may be stored as an integer or a float

# numbers

x = 1   # integer
y = 1.  # float

z = x + y

type(x), type(y), z

(int, float, 2.0)

boolean¶

bookean variables are important cause they enable conditional statement if something is true then do something, otherwise do something else

True or False

True

True == 1 # True is actually mapped to 0, False to 1. This allows to use math to write statements

True

True & False  # and

False


True | False  # or

True

String¶

# string

x = "hello world"

type(x), len(x), x[0]

(str, 11, 'h')

Data types are complex objects, they can have “methods” associated to them. The methods are invoked by a ‘.’ after the variable. They are functions that are automaticallly applied to the variable that owns them. On Colab the methods availeble will show if you put a dot after a variable. (Functions are described below)

x.

  File "<ipython-input-9-a2885923daf8>", line 1
    x.
      ^
SyntaxError: invalid syntax

x.split(" ")

['hello', 'world']

# built-in string manipulate functions
x.replace('l', 'L'), x.strip("h")

('heLLo worLd', 'ello world')

x.upper()

strings are essentiallly lists on characters and inherit some behaviors from lists (see below)

x * 3

str(1) # convert number to string

list, tuple¶

lists and tuples are python native containers: you can use them to store a set of vsariables together

# list, many built-in methods
x = [1, 'b', 2]
x.append(3)
x

[1, 'b', 2, 3]

x.pop(1) # remove item
x

[1, 2, 3]

x[2:4]

[3]

x[-2]

2

1 in x

True

# tuples, does not support item assignment, if you put things in tuples they are "safe" and cannot be changed
x = (1, 'b')
x

(1, 'b')

x[0] = 2

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-21-ebac946b3580> in <module>()
----> 1 x[0] = 2

TypeError: 'tuple' object does not support item assignment

dictionary¶

dictionaries are key-value paired containers: each element has a name by which you can retrieve it

# dictionary
dic ={'one':1, 'two':2, 'three':3}

dic.keys()

dict_keys(['one', 'two', 'three'])

dic['one']

1

# append to dict
dic['four'] = 4
dic

{'four': 4, 'one': 1, 'three': 3, 'two': 2}

conditionals and loops¶

if statement¶

x = 2
if x == 1:
    print("1")
else:
    print(' not 1')

 not 1

x = 2

if x == 1:
    print("1")
elif x == 2:
    print("2")
elif x == 3:
    print("3")
else:
    print("others")

for loop¶

for i in range(3):
    print(i)

0
1
2

# use zip to loop over multiple variables

x = ['a', 'b', 'c', ]
y = [1, 2, 3]

for i, j in zip(x, y):
    print(i, j)

a 1
b 2
c 3

# use enumerate get index of each loop

x = ['a', 'b', 'c', ]

for i, xi in enumerate(x):
        print(i, xi)

0 a
1 b
2 c

while loop¶

x = 0
while x < 3:
    print(x)
    x += 1

0
1
2

functions¶

def add(a=0, b=0):
    """
    Add two numbers
    Args:
        a: number, default 0
        b: number, default 0

    Returns:
        return the sum

    """
    result = a + b
    return result

the bit between “”" and “”" is called a “docstring” and all functions should have one to describe input, output, and puspose

add(1, 2)

3

classes¶

Creating a new class creates a new type of object, like a complex variable that has inside of itself variables and methods

class Car:
    def __init__(self,brand,**kwargs):
        self.brand = brand
        self.kwargs = kwargs

    def get_brand(self):
        return self.brand

    def get_prams(self):
        return self.kwargs

# Instantiate
newCar = Car('BMW', price=10000, color='red', model='m')

after a class is defined, an object of the class is creted by “instantiating” it, defining all required values

newCar.get_brand()

newCar.get_prams()

{'color': 'red', 'model': 'm', 'price': 10000}

Inheritance: a class can contain the initialization of another (parent) class¶

class Coupe(Car):
    def __init__(self, brand,**kwargs):
        super().__init__(brand,**kwargs) # works in phthon 3
        #super(Coupe, self).__init__(brand,**kwargs)  # works in python 2/3

    def test(self):
        print(' class works')

    def __str__(self):
        # for print
        return self.brand

# Instantiate
newCoupe = Coupe('BMW', price=10000, color='red', model='m')

newCoupe.test()

 class works

newCoupe.__str__()

standard library¶

Python comes with a standard library of modules for performing common tasks,

all built-in modules https://docs.python.org/3/library/

interact with system¶

import os

os.getcwd()  # return the current working directory
#os.chdir()  # change working dir

os.system('mkdir something') # run a terminal command

0

returns the code of the command https://man7.org/linux/man-pages/man1/mkdir.1.html

accessing files in your working folder¶

import glob

useful to list files inside of the directory in which you are working (I have no files ending in “.py” here)

glob.glob("*.py")

[]

glob.glob("*")

['today', 'something', 'sample_data']

I can also access terminal command by using ! at the beginning of the line if I am working in a notebook. Here “ls” stands for list, and lists all files in this directory

!ls

sample_data  something	today

random¶

random variables are going to be super important for us!

packages¶

from numpy import random

random.choice(['apple', 'pear', 'banana'])

random.randint(0,10) #random integer between 0 and 10

8

random.randn(2,3) # 6 numbers each a random drawin a gaussian distribution with mean 0 and standard deviation 1 organized as a 2x3 array

array([[ 0.19473473,  0.6750933 , -1.83021753],
       [-0.31215882, -0.29824092, -0.06436665]])

numpy/scipy¶

numpy is the core module for linear algebra and multidimensional arrays; scipy provides additional functions for scientific computing

import numpy as np
import scipy as sp

# create an array
a = np.array([1,2,3])
b = np.array([(1.5,2,3), (4,5,6)], dtype = np.float32)
c = np.random.random([3, 3])
c

array([[0.62274266, 0.37133858, 0.99875643],
       [0.9585066 , 0.49238662, 0.41288253],
       [0.08803071, 0.07511918, 0.76451357]])

c.size, c.shape, c.ndim

(9, (3, 3), 2)

# slicing
c[:, 2:]

array([[0.99875643],
       [0.41288253],
       [0.76451357]])

c + 1  # broadcasting

array([[1.62274266, 1.37133858, 1.99875643],
       [1.9585066 , 1.49238662, 1.41288253],
       [1.08803071, 1.07511918, 1.76451357]])

np.exp(c)

array([[1.86403345, 1.44967383, 2.71490356],
       [2.60779908, 1.6362166 , 1.5111675 ],
       [1.09202165, 1.07801263, 2.1479493 ]])

pandas¶

library for data manipulation

import pandas as pd

df = pd.DataFrame(columns=['a', 'b', 'c'])

df['a'] = ['a1', 'a2', 'a3']
df['b'] = ['b1', 'b2', 'b3']
df['c'] = ['c1', 'c2', 'c3']
df.head(3)

df.iloc[2]

a    a3
b    b3
c    c3
Name: 2, dtype: object

df.append({'a':'a4', 'b':'b4', 'c':'c4'}, ignore_index=True)

matplotlib¶

plotting figures

import matplotlib.pyplot as plt
# show figures inside notebook
%matplotlib inline

x = np.linspace(-10, 10, 100)
y = np.random.randn(100)
plt.plot(x, y, color='r')
plt.xlabel('x')
plt.ylabel('y')

resources¶

Python Cheatsheet

The Python Tutorial

Python Data Science Handbook

Hands-on Machine Learning with Scikit-Learn, Keras and TensorFlow