Adding Title and Labels

A chart without title and labels does not convey meaningful information. Matplotlib allows you to add a title and labels to the axes using the title(), xlabel(), and ylabel() functions as follows:

%matplotlib inline
import matplotlib.pyplot as plt
 
plt.plot(
    [1,2,3,4,5,6,7,8,9,10],
    [2,4.5,1,2,3.5,2,1,2,3,2]
)
plt.title("Results")    # sets the title for the chart
plt.xlabel("Semester")  # sets the label to use for the x‐axis
plt.ylabel("Grade")     # sets the label to use for the y‐axis 

Figure 4.2 shows the chart with the title, as well as the labels for the x‐ and y‐axes.

Styling

Matplotlib lets you adjust every aspect of your plot and create beautiful charts. However, it is very time consuming to create really beautiful charts and plots. To help with this, matplotlib ships with a number of predefined styles. Styles allow you to create professional‐looking charts using a predefined look‐and‐feel without requiring you to customize each element of the chart individually.

The following example uses the ggplot style, based on a popular data visualization package for the statistical programming language R:

%matplotlib inline
import matplotlib.pyplot as plt
 
from matplotlib import style
style.use("ggplot")
 
plt.plot(
    [1,2,3,4,5,6,7,8,9,10],
    [2,4.5,1,2,3.5,2,1,2,3,2]
)
plt.title("Results")     # sets the title for the chart
plt.xlabel("Semester")   # sets the label to use for the x-axis
plt.ylabel("Grade")      # sets the label to use for the y-axis 

The chart styled using ggplot is shown in Figure 4.3.

Figure 4.4 shows the same chart with the grayscale styled applied.

You can use the style.available property to see the list of styles supported:

print(style.available) 

Here is a sample output:

['seaborn-dark', 'seaborn-darkgrid', 'seaborn-ticks', 'fivethirtyeight', 'seaborn-whitegrid', 'classic', '_classic_test', 'fast', 'seaborn-talk', 'seaborn-dark-palette', 'seaborn-bright', 'seaborn-pastel', 'grayscale', 'seaborn-notebook', 'ggplot', 'seaborn-colorblind', 'seaborn-muted', 'seaborn', 'Solarize_Light2', 'seaborn-paper', 'bmh', 'seaborn-white', 'dark_background', 'seaborn-poster', 'seaborn-deep'] 

Plotting Multiple Lines in the Same Chart

You can plot multiple lines in the same chart by calling the plot() function one more time, as the following example shows:

%matplotlib inline
import matplotlib.pyplot as plt
 
from matplotlib import style
style.use("ggplot")
 
plt.plot(
    [1,2,3,4,5,6,7,8,9,10],
    [2,4.5,1,2,3.5,2,1,2,3,2]
)
 
plt.plot(
    [1,2,3,4,5,6,7,8,9,10],
    [3,4,2,5,2,4,2.5,4,3.5,3]
)
 
plt.title("Results")     # sets the title for the chart
plt.xlabel("Semester")   # sets the label to use for the x-axis
plt.ylabel("Grade")      # sets the label to use for the y-axis 

Figure 4.5 shows the chart now containing two line graphs.

Adding a Legend

As you add more lines to a chart, it becomes more important to have a way to distinguish between the lines. Here is where a legend is useful. Using the previous example, you can add a label to each line plot and then show a legend using the legend() function as follows:

%matplotlib inline
import matplotlib.pyplot as plt
 
from matplotlib import style
style.use("ggplot")
 
plt.plot(
    [1,2,3,4,5,6,7,8,9,10],
    [2,4.5,1,2,3.5,2,1,2,3,2],
    label="Jim"
)
 
plt.plot(
    [1,2,3,4,5,6,7,8,9,10],
    [3,4,2,5,2,4,2.5,4,3.5,3],
    label="Tom"
)
 
plt.title("Results")     # sets the title for the chart
plt.xlabel("Semester")   # sets the label to use for the x-axis
plt.ylabel("Grade")      # sets the label to use for the y-axis
plt.legend() 

Figure 4.6 shows the chart with a legend displayed.

Adding Another Bar to the Chart

Just like adding an additional line chart to the chart, you can add another bar graph to an existing chart. The following statements in bold do just that:

%matplotlib inline
import matplotlib.pyplot as plt
from matplotlib import style
 
style.use("ggplot")
 
plt.bar(
    [1,2,3,4,5,6,7,8,9,10],
    [2,4.5,1,2,3.5,2,1,2,3,2],
    label = "Jim",
    color = "m",                    # for magenta
    align = "center",
    alpha = 0.5
)
 
plt.bar(
    [1,2,3,4,5,6,7,8,9,10],
    [1.2,4.1,0.3,4,5.5,4.7,4.8,5.2,1,1.1],
    label = "Tim",
    color = "g",          # for green
    align = "center",
    alpha = 0.5
)
 
 
plt.title("Results")
plt.xlabel("Semester")
plt.ylabel("Grade")
 
plt.legend()
plt.grid(True, color="y") 

Because the bars might overlap each with other, it is important to be able to distinguish them by setting their alpha to 0.5 (making them translucent). Figure 4.8 shows the two bar graphs in the same chart.

Changing the Tick Marks

So far in our charts, the tick marks on the x‐axis always displays the value that was supplied (such as 2, 4, 6, and so on). But what if your x‐axis label is in the form of strings like this?

rainfall = [17,9,16,3,21,7,8,4,6,21,4,1]
months = ['Jan','Feb','Mar','Apr','May','Jun',
          'Jul','Aug','Sep','Oct','Nov','Dec'] 

In this case, you might be tempted to plot the chart directly as follows:

%matplotlib inline
import matplotlib.pyplot as plt
 
rainfall = [17,9,16,3,21,7,8,4,6,21,4,1]
months = ['Jan','Feb','Mar','Apr','May','Jun',
          'Jul','Aug','Sep','Oct','Nov','Dec']
 
plt.bar(months, rainfall, align='center', color='orange' )
plt.show() 

The preceding code snippet will create the chart shown in Figure 4.9.

Look carefully at the x‐axis: the labels have been sorted alphabetically, and hence the chart does not show the amount of rainfall from Jan to Dec in the correct order. To fix this, create a range object matching the size of the rainfall list, and use it to plot the chart. To ensure that the month labels are displayed correctly on the x‐axis, use the xticks() function:

%matplotlib inline
import matplotlib.pyplot as plt
 
rainfall = [17,9,16,3,21,7,8,4,6,21,4,1]
months = ['Jan','Feb','Mar','Apr','May','Jun',
          'Jul','Aug','Sep','Oct','Nov','Dec']
 
plt.bar(range(len(rainfall)), rainfall, align='center', color='orange' )
plt.xticks(range(len(rainfall)), months, rotation='vertical')
plt.show() 

The xticks() function sets the tick labels on the x‐axis, as well the positioning of the ticks. In this case, the labels are displayed vertically, as shown in Figure 4.10.

Exploding the Slices

The explode parameter specifies the fraction of the radius with which to offset each wedge. In the preceding example, we have set the explode parameter to all zeros:

explode     = (0,0,0,0,0,0,0,0) 

Say that we need to highlight the market share of the Firefox and Safari browsers. In that case, we could modify the explode list as follows:

explode     = (0,0,0.5,0,0.8,0,0,0) 

Refreshing the chart, you will see the two slices exploding (separating) from the main pie (see Figure 4.12).

Displaying Custom Colors

By default, matplotlib will decide on the colors to use for each of the slices in the pie chart. Sometimes the colors that are selected may not appeal to you. But you can certainly customize the chart to display using your desired colors.

You can create a list of colors and then pass it to the colors parameter:

%matplotlib inline
import matplotlib.pyplot as plt
 
labels      = ["Chrome", "Internet Explorer",
               "Firefox", "Edge","Safari",
               "Sogou Explorer","Opera","Others"]
 
marketshare = [61.64, 11.98, 11.02, 4.23, 3.79, 1.63, 1.52, 4.19]
explode     = (0,0,0.5,0,0.8,0,0,0)
colors      = ['yellowgreen', 'gold', 'lightskyblue', 'lightcoral']
 
plt.pie(marketshare,
        explode = explode,  # fraction of the radius with which to
                            # offset each wedge
        labels = labels,
        colors = colors,
        autopct="%.1f%%",   # string or function used to label the
                            # wedges with their numeric value
        shadow=True,
        startangle=45)      # rotates the start of the pie chart by
                            # angle degrees counterclockwise from the
                            # x-axis
plt.axis("equal")           # turns off the axis lines and labels
plt.title("Web Browser Marketshare - 2018")
plt.show() 

Since there are more slices than the colors you specified, the colors will be recycled. Figure 4.13 shows the pie chart with the new colors.

Rotating the Pie Chart

Observe that we have set the startangle parameter to 45. This parameter specifies the degrees by which to rotate the start of the pie chart, counterclockwise from the x‐axis. Figure 4.14 shows the effect of setting the startangle to 0 versus 45.

Displaying a Legend

Like the line and bar charts, you can also display a legend in your pie charts. But before you can do that, you need to handle the return values from the pie() function:

pie = plt.pie(marketshare,
        explode = explode,  # fraction of the radius with which to
                            # offset each wedge
        labels = labels,
        colors = colors,
        autopct="%.1f%%",   # string or function used to label the
                            # wedges with their numeric value
        shadow=True,
        startangle=45)      # rotates the start of the pie chart by
                            # angle degrees counterclockwise from the
                            # x-axis 

The pie() function returns a tuple containing the following values:

• patches: A list of matplotlib.patches.Wedge instances.
• texts: A list of the label matplotlib.text.Text instances.
• autotexts: A list of Text instances for the numeric labels. This will only be returned if the parameter autopct is not None.

To display the legend, use the legend() function as follows:

plt.axis("equal")           # turns off the axis lines and labels
plt.title("Web Browser Marketshare - 2018")
plt.legend(pie[0], labels, loc="best")
plt.show() 

Figure 4.15 shows the legend displaying on the pie chart.

The positioning of the legend can be modified through the loc parameter. It can take in either a string value or an integer value. Table 4.1 shows the various values that you can use for the loc parameter.

Saving the Chart

So far, you have been displaying the charts in a browser. At times, it is useful to be able to save the image to disk. You can do so using the savefig() function as follows:

%matplotlib inline
import matplotlib.pyplot as plt
 
labels      = ["Chrome", "Internet Explorer",
               "Firefox", "Edge","Safari",
               "Sogou Explorer","Opera","Others"]
 
…
plt.axis("equal")           # turns off the axis lines and labels
plt.title("Web Browser Marketshare - 2018")
plt.savefig("Webbrowsers.png", bbox_inches="tight")
plt.show() 

Setting the bbox_inches parameter to tight removes all of the extra white space around your figure.

Combining Plots

You can combine multiple scatter plots into one chart as follows:

%matplotlib inline
import matplotlib.pyplot as plt
 
import numpy as np
 
a = np.arange(1,4.5,0.1)   # 1.0, 1.1, 1.2, 1.3…4.4
plt.plot(a, a**2, 'y^',    # yellow triangle_up marker
         a, a**3, 'bo',    # blue circle
         a, a**4, 'r--',)  # red dashed line
 
plt.axis([0, 4.5, 0, 70])  # xmin, xmax, ymin, ymax
plt.show() 

Figure 4.17 shows the chart displaying three scatter plots. You can customize the shape of the points to draw on the scatter plot. For example, y^ indicates a yellow triangle‐up marker, bo indicates a blue circle, and so on.

Subplots

You can also plot multiple scatter plots separately and combine them into a single figure:

%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
 
a = np.arange(1,5,0.1)
 
plt.subplot(121)            # 1 row, 2 cols, chart 1
plt.plot([1,2,3,4,5],
         [1,8,27,64,125],
         'y^')
 
plt.subplot(122)            # 1 row, 2 cols, chart 2
plt.plot(a, a**2, 'y^',
         a, a**3, 'bo',
         a, a**4, 'r--',)
 
plt.axis([0, 4.5, 0, 70])   # xmin, xmax, ymin, ymax
plt.show() 

Figure 4.18 shows two charts displayed in a single figure.

The subplot() function adds a subplot to the current figure. One of the arguments it takes in has the following format: nrow,ncols,index. In the preceding example, the 121 means “1 row, 2 columns, and chart 1.” Using this format, you can have up to a maximum of nine figures. The subplot() function can also be called with the following syntax:

plt.subplot(1,2,1)            # 1 row, 2 cols, chart 1 

Using this syntax, you can now have more than 10 charts in a single figure.

Displaying Categorical Plots

The first example that you will plot is called a categorical plot (formerly known as a factorplot). It is useful in cases when you want to plot the distribution of a certain group of data. Suppose that you have a CSV file named drivinglicense.csv containing the following data:

gender,group,license
men,A,1
men,A,0
men,A,1
women,A,1
women,A,0
women,A,0
men,B,0
men,B,0
men,B,0
men,B,1
women,B,1
women,B,1
women,B,1
women,B,1 

This CSV file shows the distribution of men and women in two groups, A and B, with 1 indicating that the person has a driver's license and a 0 indicating no driver's license. If you are tasked with plotting a chart to show the proportion of men and women in each group that has a driver's license, you can use Seaborn's categorical plot.

First, import the relevant modules:

import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd 

Load the data into a Pandas dataframe:

#---load data---
data = pd.read_csv('drivinglicense.csv') 

Call the catplot() function with the following arguments:

#---plot a factorplot---
g = sns.catplot(x="gender", y="license", col="group",
        data=data, kind="bar", ci=None, aspect=1.0) 

You pass in the dataframe through the data parameter, and you specify the gender as the x‐axis. The y‐axis will tabulate the proportion of men and women who have a driver's license, and hence you set y to license. You want to separate the chart into two groups based on group, hence you set col to group.

Next, you set the labels on the chart:

#---set the labels---
g.set_axis_labels("", "Proportion with Driving license")
g.set_xticklabels(["Men", "Women"])
g.set_titles("{col_var} {col_name}")
 
#---show plot---
plt.show() 

Figure 4.19 shows the categorical plot drawn by Seaborn. As you can see, 2/3 of the men and 1/3 of the women have driver's licenses in Group A, while in Group B, 1/4 of the men and all the women have driver's licenses. Neat, isn't it?

Let's take a look at another example of catplot. Using the Titanic dataset, let's plot a chart and see what the survival rate of men, women, and children looks like in each of the three classes.

import matplotlib.pyplot as plt
import seaborn as sns
 
titanic = sns.load_dataset("titanic")
g = sns.catplot(x="who", y="survived", col="class",
        data=titanic, kind="bar", ci=None, aspect=1)
 
g.set_axis_labels("", "Survival Rate")
g.set_xticklabels(["Men", "Women", "Children"])
g.set_titles("{col_name} {col_var}")
 
#---show plot---
plt.show() 

Figure 4.20 shows the distribution of the data based on classes. As you can see, both women and children have a higher chance of survival if they are in the first‐ and second‐class cabins.

Displaying Lmplots

Another plot that is popular in Seaborn is the lmplot. An lmplot is a scatter plot. Using another built‐in dataset from Seaborn, you can plot the relationships between the petal width and petal length of an iris plant and use it to determine the type of iris plants: setosa, versicolor, or virginica.

import seaborn as sns
import matplotlib.pyplot as plt
 
#---load the iris dataset---
iris = sns.load_dataset("iris")
 
#---plot the lmplot---
sns.lmplot('petal_width', 'petal_length', data=iris,
           hue='species', palette='Set1',
           fit_reg=False, scatter_kws={"s": 70})
 
#---get the current polar axes on the current figure---
ax = plt.gca()
ax.set_title("Plotting using the Iris dataset")
 
#---show the plot---
plt.show() 

Figure 4.21 shows the scatter plot created using the lmplot() function.

Displaying Swarmplots

A swarmplot is a categorical scatterplot with nonoverlapping points. It is useful for discovering the distribution of data points in a dataset. Consider the following CSV file named salary.csv, which contains the following content:

gender,salary
men,100000
men,120000
men,119000
men,77000
men,83000
men,120000
men,125000
women,30000
women,140000
women,38000
women,45000
women,23000
women,145000
women,170000 

You want to show the distribution of salaries for men and women. In this case, a swarmplot is an ideal fit. The following code snippet does just that:

import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
 
sns.set_style("whitegrid")
 
#---load data---
data = pd.read_csv('salary.csv')
 
#---plot the swarm plot---
sns.swarmplot(x="gender", y="salary", data=data)
 
ax = plt.gca()
ax.set_title("Salary distribution")
 
#---show plot---
plt.show() 

Figure 4.22 shows that, in this group, even though women have the highest salary, it also has the widest income disparity.