Python Pandas - Visualization

Visualization of data plays an important role in data analysis, it helps you represent the data graphically for better understanding, and identifying the patterns. However, Pandas library is primarily used for data manipulation and analysis but it also provides the data visualization capabilities by using the Python's Matplotlib library support.

In Python, the Pandas library provides a basic method called .plot() for generating a wide variety of visualizations along the different specialized plotting methods. These visualizations tools are built on top of the Python's Matplotlib library, offering flexibility and customization options.

Behind the scenes, every plot generated by Pandas is actually a Matplotlib object. This integration allows users to leverage Matplotlib's extensive customization options for fine-tuning Pandas-generated plots.

In this tutorial, we will learn about basics of visualizing data using the Pandas data structures.

Setting Up the Environment for Visualization

Before learning about Pandas data Visualization, we should ensure that Matplotlib library is installed. Following is the command for installing the Matplotlib library −

pip install pandas matplotlib

Importing Libraries

Along with the import pandas as pd you need to import the Matplotlib's functional interface for displaying, customizing, and saving plots using the following command −

import matplotlib.pyplot as plt

Displaying the plots

In environments like Jupyter Notebook or IPython shell, plots are often displayed automatically as they are generated. However, in a standard Python script or shell, this does not happen automatically. To explicitly display a plot in such environments, we need to call the following command −

plt.show()

This command renders the Matplotlib figure object in a GUI window.

Pandas Basic Plotting Method

The Pandas library provides a basic plotting method called plot() on both the Series and DataFrame objects for plotting different kind plots. This method is a simple wrapper around the matplotlib plt.plot() method.

Syntax

Following is the syntax of the Pandas .plot() method −

DataFrame.plot(*args, **kwargs)

Where,

• kind: Specifies the type of plot (default: 'line').

• *args:

• **kwargs:

Example

Here is the following example of plotting a random DataFrame data using the Pandas plot() method.

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.randn(10,4),index=pd.date_range('1/1/2000',
   periods=10), columns=list('ABCD'))

# Plotting the DataFrame
df.plot()
plt.show()

Its output is as follows −

If the index consists of dates, then the pandas .plot() method calls the Matplotlib gct().autofmt_xdate() to format the x-axis labels.

Also we can plot one column versus another using the x and y keywords.

Types of Plots Available in Pandas

Pandas supports various plot types through the kind parameter or specialized plotting methods. Following is the overview of the different plotting methods −

Example: Plotting Bar Plot with plot() method

Let us now see what a Bar Plot is by creating one. A bar plot can be created in the following way −

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.rand(10,4), columns=['a','b','c','d'])

# Plotting the bar plot
df.plot(kind='bar')
plt.show()

Its output is as follows −

To produce a stacked bar plot, pass stacked=True −

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.rand(10,4), columns=['a','b','c','d'])

# Plotting the stacked Bar plot
df.plot(kind='bar', stacked=True)
plt.show()

Its output is as follows −

To get horizontal bar plots, use the barh option −

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.rand(10,4), columns=['a','b','c','d'])

# Plotting the horizontal bar plot
df.plot(kind='barh', stacked=True)
plt.show()

Its output is as follows −

Histograms

Histograms can be plotted using the hist option of the plot() method kind argument. We can specify number of bins.

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

# Creating a random DataFrame 
df = pd.DataFrame({'a':np.random.randn(1000)+1,'b':np.random.randn(1000),
'c':np.random.randn(1000) - 1}, columns=['a', 'b', 'c'])

df.plot(kind='hist', bins=20)
plt.show()

Its output is as follows −

Box Plots

Box plot can be drawn calling 'box' option for both the Series and DataFrame objects to visualize the distribution of values within each column.

For instance, here is a boxplot representing five trials of 10 observations of a uniform random variable on [0,1).

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.rand(10, 5), columns=['A', 'B', 'C', 'D', 'E'])

df.plot(kind='box')
plt.show()

Its output is as follows −

Area Plot

Area plot can be created using the plot(kind='area') option.

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.rand(10, 4), columns=['a', 'b', 'c', 'd'])
df.plot(kind='area')

plt.show()

Its output is as follows −

Scatter Plot

Scatter plot can be created using the plot(kind='scatter') option.

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

# Creating a random DataFrame 
df = pd.DataFrame(np.random.rand(50, 4), columns=['a', 'b', 'c', 'd'])
df.plot(kind='scatter', x='a', y='b')

plt.show()

Its output is as follows −

Pie Chart

Pie chart can be created using the plot(kind='pie') option.

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

# Creating a random DataFrame 
df = pd.DataFrame(3 * np.random.rand(4), index=['a', 'b', 'c', 'd'], columns=['x'])
df.plot(kind='pie', subplots=True)

plt.show()

Its output is as follows −