Graphical Data Representation with Python – Matplotlib: Part Two

In the previous article, I talked about matplotlib for drawing generic curves; in this one, I will show how to draw scatter plots, histograms, and time series.

I remind you that for the following examples, I will use Jupyter notebook. Therefore, if you’re not familiar with it, I invite you to take a look at this article.

Creating the Jupyter Notebook Project

Open a shell, create a new directory, and run the command jupyter notebook

mkdir EXAMPLE
cd EXAMPLE/
jupyter notebook

A browser window will open.

At this point, select the Python 3 interpreter to initialize the environment.

Scatter Plots

First, let’s import the necessary libraries:

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

To generate points in the scatter space, we use numpy.

#Generating random points
randomPoints=np.random.randn(100,2);

np.random.randn returns samples in the scatter space. If positive integer arguments are provided, randn generates an array (d0, d1, …, dn), filled with random floats sampled from a “normal” (Gaussian) distribution with a mean of 0 and variance of 1. If no arguments are provided, a single random float sample from the standard distribution will be returned.

The previous instruction will generate 100 random points. The result will look something like this:

array([[-0.28393081,  1.69444015],
       [ 1.18710769,  0.75168393],
       [-1.08797599,  0.16593107],
       [ 0.34718794,  1.50584947],
       [ 0.59907203,  0.34574815],
       [-1.17150305,  1.32782198],
       [-0.66961386,  0.13941936],
       [ 0.96038988, -0.04362326],
       [-0.42828503, -1.35202516],
       [-0.57520301, -0.5182515 ]])

At this point, we can create the dataframe that we will use as a parameter for matplotlib’s pyplot.

df = pd.DataFrame(randomPoints,columns=['x', 'y'])

And print the scatter plot:

plt.plot(df['x'],df['y'],'.')

 

 

Complete Example

If we wanted to represent more prominent points, we could use the “o” print option instead of “.”

plt.plot(df['x'],df['y'],'o')

Other options can allow us to create more meaningful plots, defining the size of the points and the color, but to do this, we will need to use the scatter function instead of plot.

Note: For our example, we will also use random data for the colors and the size of the points;

random_color=np.random.randn(100)
random_point_size=1000*np.random.randn(100)
plt.scatter(df['x'],df['y'],s=random_point_size,alpha=0.6,c=random_color)

 

<< Data Representation with Python – Matplotlib