A picture is worth a thousand words, and the visualization of data plays a critical role in finding hidden patterns in the data. Over a period of time, a variety of graphs have been developed to represent different relationships between different types of variables. In this chapter, we will see how to use these different graphs in different contexts and how to plot them using Matplotlib.

The line plot is used to represent a relationship between two continuous variables. It is typically used to represent the trend of a variable over time, such as GDP growth rate, inflation, interest rates, and stock prices over quarters and years. All the graphs we have seen in Chapter 1, Anatomy of Matplotlib are examples of a line plot.

We will use the Google Stock Price data for plotting time series line plot. We have the data (date and daily closing price, separated by commas) in a .csv file without a header, so we will use the pandas library to read it and pass it on to the matplotlib.pyplot function to plot the graph.

Let's now import required libraries with the following code:

import matplotlib...

Bar plots are the graphs that use bars to compare different categories of data. Bars can be shown vertically or horizontally, based on which axis is used for a categorical variable. Let's assume that we have data on the number of ice creams sold every month in an ice cream parlor over a period of one year. We can visualize this using a bar plot.

We will use the Python calendar package to map numeric months (1 to 12) to the corresponding descriptive months (January to December).

Before we plot the graph, we need to import the necessary packages:

import matplotlib.pyplot as plt
import numpy as np
import calendar

A scatter plot is used to compare distribution of two variables and see whether there is any correlation between them. If there are distinct clusters/segments within the data, it will be clear in the scatter plot.

Import the following libraries:

import matplotlib.pyplot as plt
import pandas as pd

We will use pandas to read the Excel files.

The following code block draws a scatter plot that depicts the relationship between the age and the weight of people:

1. Set the figure size (width and height) to (10, 6) inches:

plt.figure(figsize...

A bubble plot is drawn using the same plt.scatter() method. It is a manifestation of the scatter plot, where each point on the graph is shown as a bubble. Each of these bubbles can be displayed with a different color, size, and appearance.

Import the required libraries. We will use pandas to read the Excel file:

import matplotlib.pyplot as plt
import pandas as pd

The following code block plots a bubble plot, for which we have seen a scatter plot earlier:

1. Load the Iris dataset:

iris = pd.read_csv('iris_dataset.csv',...

A stacked plot represents the area under the line plot, and multiple line plots are stacked one over the other. It is used to provide a visualization of the cumulative effect of multiple variables being plotted on the y axis.

We will plot the number of product defects by a defect reason code, for three months stacked together to give a cumulative picture for the quarter.

Import the required libraries:

import numpy as np
import matplotlib.pyplot as plt

The following are the steps to plot a stacked plot:

1. Define the data for the...

A pie plot is used to represent the contribution of various categories/groups to the total. For example, the contribution of each state to the national GDP, contribution of a movie to the total number of movies released in a year, student grades (A, B, C, D, and E) as a percentage of the total class size, or a distribution of monthly household expenditure on groceries, vegetables, utilities, apparel, education, healthcare, and so on.

Import the required library:

import matplotlib.pyplot as plt

The following code block plots a pie...

A table chart is a combination of a bar chart and a table, representing the same data. So, it is a combination of pictorial representation along with the corresponding data in a table.

We will consider an example of the number of batteries sold in each year, with different ampere hour (Ah) ratings. There are two categorical variables: year and Ah ratings, and one numeric variable: the number of batteries sold.

Import the required libraries:

import numpy as np
import matplotlib.pyplot as plt

The following code block draws a table...

The polar plot is a chart plotted on the polar axis, which has coordinates as angle (in degrees) and radius, as opposed to the Cartesian system of x and y coordinates. We will take the example of planned versus the actual expense incurred by various departments in an organization. This is also known as a spider web plot.

Import the required libraries:

import numpy as np
import matplotlib.pyplot as plt

The histogram plot is used to draw the distribution of a continuous variable. Continuous variable values are split into the required number of bins and plotted on the x axis, and the count of values that fall in each of the bins is plotted on the y axis. On the y axis, instead of count, we can also plot percentage of total, in which case it represents probability distribution. This plot is typically used in statistical analysis.

We will use the example of data on prior work experience of participants in a lateral training program. Experience is measured in number of years.

Import the required libraries:

import matplotlib.pyplot as plt
import numpy as np

The box plot is used to visualize the descriptive statistics of a continuous variable. It visually shows the first and third quartile, median (mean), and whiskers at 1.5 times the Inter Quartile Range (IQR)—the difference between the third and first quartiles, above which are outliers. The first quartile (the bottom of rectangular box) marks a point below which 25% of the total points fall. The third quartile (the top of rectangular box) marks a point below which 75% of the points fall.

If there are no outliers, then whiskers will show min and max values.

This is again used in statistical analysis.

We will use an example of wine quality dataset for this example. We will consider three attributes...

The violin plot is a combination of histogram and box plot. It gives information on the complete distribution of data, along with mean/median, min, and max values.

We will use the same data that we used for the box plot for this example also.

Import the required libraries:

import matplotlib.pyplot as plt
import pandas as pd

The following is the code block that draws violin plots:

1. Read the data from a CSV file into a pandas DataFrame:

wine_quality = pd.read_csv('winequality.csv', delimiter=';')

2. Prepare the...

Matplotlib.pyplot has features that enable us to read .jpeg and .png images and covert them to pixel format to display as images.

Import the required library:

 import matplotlib.pyplot as plt

Here is the code block that reads a JPEG file as a pixel-valued list and displays it as an image on the screen:

1. Read the image louvre.jpg into a three-dimensional array (color images have three channels, whereas black and white images have one channel only):

image = plt.imread('louvre.jpg')

2. Print the dimensions...

A heatmap is used to visualize data range in different colors with varying intensity. Here, we take the example of plotting a correlation matrix as a heatmap. Elements of the correlation matrix indicate the strength of a linear relationship between the two variables, and the matrix contains such values for all combinations of the attributes in the given data. If the data has five attributes, then the correlation matrix will be a 5 x 5 matrix.

We will use the Wine Quality dataset for this example also. It has 12 different attributes. We will first get a correlation matrix and then plot it as a heatmap. There is no heatmap function/method as such, so we will use the same imshow method that we used to read...

The Hinton diagram is a 2D plot for visualizing weight matrices in deep-learning applications. Matplotlib does not have a direct method to plot this diagram. So, we will have to write code to plot this. The weight matrix for this is taken from one of the machine learning algorithms that classifies images.

Import the required libraries:

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

The following code block defines the function and makes a call to the function to plot the Hinton diagram:

1. Read the weight...

The contour plot is typically used to display how the error varies with varying coefficients that are being optimized in a machine learning algorithm, such as linear regression. If the linear regression coefficients are theta0 and theta1, and the error between the predicted value and the actual value is a Loss, then for a given Loss value, all the values of theta0 and theta1 form a contour. For different values of Loss, different contours are formed by varying the values of theta0 and theta1.

The data for Loss, theta0, theta1, and theta (optimal values of theta0 and theta1 that give the lowest error) are taken from one of the regression problems for plotting the contour plot. If theta0 and theta1...

Triangulations are used to plot geographical maps, which help with understanding the relative distance between various points. The longitude and latitude values are used as x, y coordinates to plot the points. To draw a triangle, three points are required; these are specified with the corresponding indices of the points on the plot. For a given set of coordinates, Matplotlib can compute triangles automatically and plot the graph, or optionally we can also provide triangles as an argument.

Import the required libraries. We will introduce the tri package for triangulations:

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.tri as tri

Stream plots, also known as streamline plots are used to visualize vector fields. They are mostly used in the engineering and scientific communities. They use vectors and their velocities as a function of base vectors to draw these plots.

Import the required libraries:

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec

The following code block creates a stream plot:

1. Prepare the data for the stream plot:

x, y = np.linspace(-3,3,100), np.linspace(-2,4,50)

2. Create the mesh grid:

X, Y = np.meshgrid...

Path is a method provided by Matplotlib for drawing custom charts. It uses a helper function patch that is provided by Matplotlib. Let's see how this can be used to draw a simple plot.

Import the required libraries. Two new packages, Path and patches, will be introduced here:

import matplotlib.pyplot as plt
from matplotlib.path import Path
import matplotlib.patches as patches

The following code block defines points and associated lines and curves to be drawn to form the overall picture:

1. Define the points along with the first curve...