Plotting with Matplotlib.pyplot

• 1. Basic Plots
  • 1.1 Installing the library
  • 1.2 Histograms
    • a. Adding color
    • b. Adding axis labels
    • c. Changing the number of bins
    • d. Adding grid to the plot
  • 1.3 Boxplots
    • a. Horizontal boxplot
    • b. Add variable name & axis label
    • c. Multiple boxplots in one plot
  • 1.4 Scatterplots
• 2. Matplotlib Inside pandas
  • 2.1 Histograms
  • 2.2 Boxplots
  • 2.3 Scatterplots

---

1. Basic Plots

• Matplotlib is considered by many as the most basic plotting library in Python.
• It offers both static and interactive visualizations in Python.
• Plotting functions in libraries such as as pandas are built on top of Matplotlib making it very fundamental for data scientists programming in Python.
• In this lecture, we will mainly look at Pyplot, a sub-library within Matplotlib consisting of all the basic plotting functions.

1.1 Installing the library

• The first step is to install the Matplotlib library.
• Run the following command in your command line prompt:

conda install matplotlib

• Depending on how you installed Python, you might need to try the following code instead (if the previous one doesn't work):

pip install matplotlib

1.2 Histograms

• Before we can call the plotting functions, we need to import the library to our current working environment (kernel):

import matplotlib.pyplot as plt

• In this example, we will take a look at the famous Old Faithful Geyser Dataset.
  • This dataset contains the waiting time between eruptions and the duration of the eruption for the Old Faithful geyser in Yellowstone National Park, Wyoming, USA.
  • There are 2 columns:
    • 'eruptions': eruption time (in mins)
    • 'waiting': waiting time to next eruption (in mins)

import pandas as pd

faithful = pd.read_csv('https://stat430.hknguyen.org/files/datasets/faithful.csv')
faithful.head()

	eruptions	waiting
0	3.600	79
1	1.800	54
2	3.333	74
3	2.283	62
4	4.533	85

# histogram of waiting time
plt.hist(faithful['waiting'])
plt.show()

a. Adding color

• Let's change the color of the plot!

plt.hist(faithful['waiting'], color='darkorange')
plt.show()

• So what are the colors available to choose?
  • Take a look at this documentation!

b. Adding axis labels

• Plots do not make sense without axis labels!
• To add an x-axis label, use plt.xlabel():

plt.hist(faithful['waiting'], color='darkorange')
plt.xlabel('Waiting time to the next eruption (in mins)')
plt.show()

• Similarly, use plt.ylabel() to add a y-axis label:

plt.hist(faithful['waiting'], color='darkorange')
plt.xlabel('Waiting time to the next eruption (in mins)')
plt.ylabel('Count')
plt.show()

c. Changing the number of bins

• Histogram can look very different depending on the number of bins used to plot the histogram!

plt.hist(faithful['waiting'], color='darkorange', bins=50)
plt.show()

plt.hist(faithful['waiting'], color='darkorange', bins=5)
plt.show()

d. Adding grid to the plot

• Sometimes, adding a background grid makes it a lot easier to "read" the plot.

plt.hist(faithful['waiting'], color='darkorange')
plt.xlabel('Waiting time to the next eruption (in mins)')
plt.ylabel('Count')
plt.grid(color='lightgrey', linewidth=0.5)
plt.show()

1.3 Boxplots

• Use plt.boxplot() to plot a boxplot:

# Boxplot of waiting time
plt.boxplot(faithful['waiting'])
plt.show()

a. Horizontal boxplot

• What’s if you want the boxplot to be horizontal instead?

plt.boxplot(faithful['waiting'], vert=False)
plt.show()

b. Add variable name & axis label

• Just like with histogram, we can add axis label to boxplot! It might also be a good idea to add variable name (also called label) to the boxplot.

plt.boxplot(faithful['waiting'], labels=['waiting'])
plt.ylabel('Time (in mins)')
plt.grid(color='lightgrey', linewidth=0.5)
plt.show()

c. Multiple boxplots in one plot

• It's often the case that you want to plot boxplots for multiple variables in the dataset.
• For example, let's examine the famous Iris Dataset.
  • This dataset contains measurements of 3 different Iris species: Setosa, Versicolor, and Virginica.
  • There are 5 columns:
    • 'Sepal.Length': the sepal length in cm.
    • 'Sepal.Width': the sepal width in cm.
    • 'Petal.Length': the petal length in cm.
    • 'Petal.Width': the petal width in cm.
    • 'Species': the specific Iris specie ('setosa', 'versicolor', 'virginica').

iris = pd.read_csv('https://stat430.hknguyen.org/files/datasets/iris.csv')
iris

	Sepal.Length	Sepal.Width	Petal.Length	Petal.Width	Species
0	5.1	3.5	1.4	0.2	setosa
1	4.9	3.0	1.4	0.2	setosa
2	4.7	3.2	1.3	0.2	setosa
3	4.6	3.1	1.5	0.2	setosa
4	5.0	3.6	1.4	0.2	setosa
...	...	...	...	...	...
145	6.7	3.0	5.2	2.3	virginica
146	6.3	2.5	5.0	1.9	virginica
147	6.5	3.0	5.2	2.0	virginica
148	6.2	3.4	5.4	2.3	virginica
149	5.9	3.0	5.1	1.8	virginica

150 rows × 5 columns

plt.boxplot([iris['Sepal.Length'], iris['Sepal.Width'], iris['Petal.Length'], iris['Petal.Width']])
plt.grid(color='lightgrey', linewidth=0.5)
plt.show()

plt.boxplot([iris['Sepal.Length'], iris['Sepal.Width'], iris['Petal.Length'], iris['Petal.Width']],
           labels=['Sepal Length', 'Sepal Width', 'Petal Length', 'Petal Width'])
plt.ylabel('Measurement (in cm)')
plt.grid(color='lightgrey', linewidth=0.5)
plt.show()

• This is clearly not very convenient! Later, we will discuss how to use the boxplot() function provided by pandas which improves the syntax significantly.

1.4 Scatterplots

• Scatterplot is one of the most important plots in Statistics! We plot scatterplot in Matplotlib using plt.scatter():

plt.scatter(x=iris['Sepal.Length'], y=iris['Sepal.Width'])
plt.show()

• Now, let’s add axis labels, change the color, add plot title and a background grid!

plt.scatter(x=iris['Sepal.Length'], y=iris['Sepal.Width'], color='darkorange')
plt.xlabel('Sepal Length (in cm)')
plt.ylabel('Sepal Width (in cm)')
plt.title('Iris Sepal Length vs. Sepal Width')
plt.grid(color='lightgrey', linewidth=0.5)
plt.show()

---

2. Matplotlib Inside pandas

• As discussed earlier, the plotting functions provided in the pandas library are built upon the functions provided by the Matplotlib library.
• You will soon find that this is much easier when you're dealing with data stored in a DataFrame (which is 95% of the time what we deal with).

2.1 Histograms

faithful['waiting'].plot.hist()

<AxesSubplot:ylabel='Frequency'>

faithful['waiting'].hist()

<AxesSubplot:>

• Both function calls above essentially do the exact the same thing.
• Now, pandas plotting functions is extremely useful when we want to layer out plots (histograms in this case).
• But do note that it plots each column of the DataFrame as its own histogram.
• In case of the Iris dataset, we will have to do some data manipulation in order for it to plot a histogram for each specie.

iris.plot.hist(alpha=0.5)

<AxesSubplot:ylabel='Frequency'>

2.2 Boxplots

• Similar to histogram, we can use plot.box() or boxplot() to plot boxplot(s) of column(s) of a DataFrame.

iris.plot.box()

<AxesSubplot:>

iris.boxplot()

<AxesSubplot:>

2.3 Scatterplots

# Scatterplot
iris.plot.scatter(x='Sepal.Length', y='Sepal.Width')

<AxesSubplot:xlabel='Sepal.Length', ylabel='Sepal.Width'>

• Alternatively, you can call function plot() and set the kind keyword to be 'scatter' for scatter plots.

iris.plot(x='Sepal.Length', y='Sepal.Width', kind='scatter')

<AxesSubplot:xlabel='Sepal.Length', ylabel='Sepal.Width'>

• We can modify the plot just as above if we import Matplotlib.pyplot and use the functions provided by Pyplot.

iris.plot.scatter(x='Sepal.Length', y='Sepal.Width', color='darkorange')
plt.xlabel('Sepal Length (in cm)')
plt.ylabel('Sepal Width (in cm)')
plt.title('Iris Sepal Length vs. Sepal Width')
plt.grid(color='lightgrey', linewidth=0.5)
plt.show()