Reading Python

A name holds a value, and = puts it there

An assignment such as count = 3 stores the value 3 under the name count. The = sign means "store", not "is equal to". So count = count + 1 is not a contradiction. It reads the old value, adds 1, and stores the result under the same name.

Values come in a few kinds. Whole numbers (3), decimal numbers (2.5), text in quotes ("cube", called a string), and True or False. The function print shows values on the screen. Text after # is a comment: a note for people that Python ignores.

names.py

count = 3
count = count + 1
name = "cube"
print(name, count)

cube 4

Arithmetic uses +, -, * (times), / (divide) and ** (power: 2 ** 3 is 8). The shortcut total += x means total = total + x.

A list keeps values in order, numbered from 0

A list is several values in square brackets, separated by commas. Each value has a position, called its index. Python counts positions from 0, so the first value is at index 0. A negative index counts from the end: -1 is the last value.

A slice such as xs[1:3] picks a run of values. It starts at the first number and stops just before the second. Leave a number out to go to the edge: xs[:2] is the first two values, and xs[2:] is everything from index 2 on. The function len gives a list's length.

lists.py

xs = [4.0, 7.5, 1.0, 3.0]
print(xs[0], xs[2], xs[-1], len(xs))
print(xs[1:3])

4.0 1.0 3.0 4
[7.5, 1.0]

A for-loop repeats the indented lines once for each item

A for-loop takes the items of a list one at a time. On each pass, the name after for holds the current item, and Python runs the indented lines below. The indentation, the spaces at the start of a line, is how Python knows which lines belong to the loop. The first line that is not indented runs once, after the loop ends.

Before you read the output, predict the last two lines it prints.

loop.py

total = 0
for x in [4.0, 7.5, 1.0]:
    total = total + x
    print("added", x, "total is now", total)
print("done")

added 4.0 total is now 4.0
added 7.5 total is now 11.5
added 1.0 total is now 12.5
done

range(n) gives the whole numbers 0, 1, …, n−1, so for i in range(3): runs three times with i equal to 0, 1 and 2. An if line works the same way: its indented lines run only when its test is true. Tests use == (equal), != (not equal), < and >.

A function gives a few lines a name, so you can run them again

A function is a named block of lines. def starts one. The names in parentheses are its arguments: the values you hand it. return sends a value back and ends the function. You call a function by writing its name followed by values in parentheses. In a def line, an argument written as label="value" gets a default: the value it takes when a call leaves it out. A call may also give its own value by name.

functions.py

def square(x):
    return x * x

def describe(x, label="value"):
    if x > 10:
        return label + " is big"
    return label + " is small"

print(square(3), square(2.5))
print(describe(30), describe(4, label="score"))

9 6.25
value is big score is small

describe(30) leaves label out, so it uses the default, "value". describe(4, label="score") gives its own value, so the second answer starts with "score".

import brings in code that someone else wrote

A library is a collection of functions that you can load into your program. import math loads the standard math library, and math.sqrt is its square-root function. The dot means "the thing inside". import numpy as np loads NumPy, the library for arrays of numbers, under the short name np. Units after this one use NumPy and PyTorch, a similar library for machine learning.

A dot after a value works the same way: pts.shape reads a property that the value carries, and pts.sum() calls a function that belongs to it.

Square brackets with a comma pick rows and columns of a table

A list can hold lists. Then it works like a table: grid[1] is the second row, and grid[1][0] is the first value in that row.

NumPy arrays let you write both indexes in one pair of brackets, separated by a comma: rows first, then columns. A bare : means "every one". So pts[:, 0] means "every row, column 0": the first number of every row. You will see this pattern often. Arrays in NumPy explains it fully.

Before you read the output, predict what pts[:, 0] prints.

tables.py

import math
import numpy as np

print(math.sqrt(16))
pts = np.array([[1, 2, 3],
                [4, 5, 6]])
print(pts.shape)
print(pts[1])
print(pts[:, 0])

4.0
(2, 3)
[4 5 6]
[1 4]

The shape (2, 3) means 2 rows and 3 columns.

A few more pieces appear in later snippets