NumPy, SciPy, and Matplotlib Primer¶

NumPy, SciPy, and Matplotlib are three complementary and important Python libraries that are useful for exploratory data analysis and machine learning. This primer is a high-level introduction to each library, providing short examples for each.

NumPy¶

NumPy provides a multidimensional array object for very fast operations on numerical data (e.g. arithmetic, logical, sorting, selecting, transforming, statistical, etc.).

Much faster than doing similar operations with normal Python data types because it uses pre-complied, vectorized C code behind the scenes.

import numpy as np

a = np.arange(15).reshape(3, 5)

print(a)
print(a.shape)
print(a.ndim)

### Output:

[[ 0,  1,  2,  3,  4],
 [ 5,  6,  7,  8,  9],
 [10, 11, 12, 13, 14]]
(3, 5)
2

A few very easy array operation examples:

import numpy as np

a = np.arange(15).reshape(3, 5)

print(a.max())
print(a.min())
print(a.sum())
print(a*2)

### Output:

14
0
105
[[ 0  2  4  6  8]
 [10 12 14 16 18]
 [20 22 24 26 28]]

SciPy¶

SciPy provides mathematical algorithms and convenience functions for working with NumPy arrays.

It comes with a host of high level commands for different scientific computing domains, including clustering, Fourier transforms, integration, interpolation, linear algebra, signal processing, among others.

The following short code block defines a function to be integrated (called the “integrand”) as f(x) = ax^2 + b. The SciPy quad function evaluates the integral over a given range, in this case from 0 to 1. The output is a tuple where the first item is the value of the integral, and the second item is the estimated error.

from scipy.integrate import quad
def integrand(x, a, b):
    return a*x**2 + b

a = 2
b = 1
I = quad(integrand, 0, 1, args=(a,b))
print(I)

### Output:

(1.6666666666666667, 1.8503717077085944e-14)

Matplotlib¶

Matplotlib provides utilities for creating static, animated, and interactive visualizations of data.

You can use it to create many different types of plots (line, histogram, scatter, countour, 3D, etc.) with full control over all labels, colors, styles, etc. Look online for examples of the kind of plot you want to make, and undoubtedly there will be some Matplotlib samples available.

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0, 2*np.pi, 50)
plt.plot(x, np.sin(x))
plt.show()

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0, 2*np.pi, 50)
plt.plot(x, np.sin(x), 'r-o', x, np.sin(2*x), 'g--')
plt.show()

# save the image to file instead
plt.savefig('my_sinwave.png')

NumPy, SciPy, and Matplotlib Primer¶

NumPy¶

SciPy¶

Matplotlib¶

Additional Resources¶