packages.md

Packages and modules lab

The goal of this lab is to get familiar importing existing packages/modules and to create and import your own modules.

You should retype these exercises rather than cutting/pasting, which doesn't help you remember anything.

Importing standard packages

By now, you've probably seen lots of examples and have used import many times for various projects. Just for completeness, let's go through the basics.

In a file called basics.py import the math, numpy, and pandas packages with the following.

import math
import numpy as np
import pandas as pd

Please note the standard aliases given for the second two packages; this is what you will see in all code examples and it's good to keep to the standard. The as np is an alias that we can use as a shorthand.

Let's verify that the import statements have worked:

print(math.pi)
print(np.sum(np.array([1,2,3])))

df = pd.DataFrame([[100,'parrt'],[105,'tombu']],
                  columns=['ID','user'])
print(df)

The output you get should look like this:

$ python basics.py 
3.141592653589793
6
    ID   user
0  100  parrt
1  105  tombu

Now, let's try out a variation on the import statement that imports a single variable instead of all variables. It allows us to directly reference pi instead of the package name prefix math.pi. In the same file or different file, execute this:

from math import pi
print(pi)

Importing our own module

To learn more about how modules work, let's create our own module. The good news is it's just another Python file. If we import it into another file it acts like a module rather than a script that we execute. As a general rule, modules that you import should not have a main program (code outside of functions). The only exception is when you need to create a variable that you want to be available to the file importing your module.

Create a file called mymath.py that will hold a variable and function definition mimicking standard math stuff:

pi = 3.14159

def pow2(n):
    v = 1
    for i in range(n):
        v *= 2
    return v

Now, create a script called usemymath.py that will import that code and test it:

from mymath import *

print(pi)
print(pow2(0))
print(pow2(10))

From the command line the output looks like this:

$ python usemymath.py 
3.14159
1
1024

The whole point of this exercise is to learn that organizing related functions into modules that you can reuse makes you more productive. You can even publish these packages so that other people can use your code.

If you get stuck, you can check out the solution.