Programming in PythonPackages

A package is a collection of Python files that provide functionality beyond the core functionality available in every Python program. Packages achieve separation of concerns at the community level: someone else solves a problem of general interest, and then you can leverage their work and focus on applying it to the problem at hand.

Many Python packages are available in every standard distribution of Python and can be used without having to worry about whether they're installed. These packages make up the standard library. To see a list of standard library packages, visit the standard library page of the Python documentation. Here's an example showing how to import the math package and use the sqrt function it contains:

import math
math.sqrt(3)

Note that we access names like sqrt provided by the package using the dot syntax math.sqrt. This is common practice, and it's a good idea because if functions are called in a way that makes it clear what package they came from, then (1) you can use the same name in multiple packages, and (2) you can easily identify which package that is supplying each function. We can also import individual functions and skip the dot syntax:

from math import sqrt
sqrt(3)

Sometimes a package contains a subpackage which must itself be accessed with dot syntax:

from numpy.random import standard_normal
standard_normal()

Continue

Scientific computing packages in Python

Here are some of the most important scientific computing packages (along with very brief code snippets to give you a sense of what calling the packages looks like in practice):

NumPy. Provides multi-dimensional arrays (like vectors, matrices, and higher-order arrays).

import numpy as np
np.random.standard_normal((5,5)) # randomly fill a 5 × 5 matrix
np.full((3,3),7) # make a 3 × 3 matrix full of 7's

Note that we import numpy with the alias np for brevity.

Pandas. Provides support for tabular data.

import pandas as pd
iris = pd.read_csv("http://bit.ly/iris-dataset")
iris

SciPy. Provides scientific computing tools for optimization, numerical integration, linear algebra, statistics, etc.

from scipy.optimize import minimize
minimize(lambda x: x*(x-1), 1.0) # start from 1 and minimize x(x-1)

Matplotlib. Standard plotting package in Python. (Note: run the cell below twice to get the graph to display.)

import matplotlib.pyplot as plt
import numpy as np
plt.plot(np.cumsum(np.random.standard_normal(1000)))

SymPy. Pure math tools like symbolic integration/differentiation, number theory, etc.

from sympy import symbols, Eq, solve
x = symbols("x")
y = symbols("y")
solve([Eq(x + 5*y, 2), Eq(-3*x + 6*y, 15)], [x, y])

The example above solves the system of equations:

for and .

Exercises

Solution. from numpy import arcsin

Continue

Solution import sympy as sp

Continue

Solution import itertools

Continue