Plotting directly

with Matplotlib
Estimated time needed: 15 minutes

Objectives
After completing this lab you will be able to:

Create and customize basic plots directly with Matplotlib on dataset

Table of Contents

1. Import Libraries
2. Fetching Data
3. Line Plot
4. Scatter Plot
5. Bar Plot
6. Histogram
7. Pie
 8. Subplotting

Import Libraries
Import the matplotlib library.

In [1]: #Import Primary Modules:

import numpy as np # useful for many scientific computing in Python
import pandas as pd # primary data structure library

# use the inline backend to generate the plots within the browser
%matplotlib inline

import matplotlib as mpl

import matplotlib.pyplot as plt

# check for latest version of Matplotlib

print('Matplotlib version: ', mpl.__version__) # >= 2.0.0

Matplotlib version: 3.5.3

Fetching Data
Dataset: Immigration to Canada from 1980 to 2013 - International migration flows to and
from selected countries - The 2015 revision from United Nation's website.
In this lab, we will focus on the Canadian Immigration data and use the already cleaned
dataset and can be fetched from here.

You can refer to the lab on data pre-processing wherein this dataset is cleaned for a quick
refresh your Panads skill Data pre-processing with Pandas

In [2]: URL = "https://cf-courses-data.s3.us.cloud-object-storage.appdomain.cloud/IBMDevelo

df_can = pd.read_csv(URL)
print('Data read into a pandas dataframe!')

Data read into a pandas dataframe!

Let's take a look at the first five items in our dataset.

In [3]: df_can.head()
Out[3]: Country Continent Region DevName 1980 1981 1982 1983 1984 1985 ...

Southern Developing
0 Afghanistan Asia 16 39 39 47 71 340 ...
Asia regions

Southern Developed
1 Albania Europe 1 0 0 0 0 0 ...
Europe regions

Northern Developing
2 Algeria Africa 80 67 71 69 63 44 ...
Africa regions

American Developing
3 Oceania Polynesia 0 1 0 0 0 0 ...
Samoa regions

Southern Developed
4 Andorra Europe 0 0 0 0 0 0 ...
Europe regions

5 rows × 39 columns

Let's find out how many entries there are in our dataset.

In [4]: # print the dimensions of the dataframe

print(df_can.shape)

(195, 39)

Set the country name as index - useful for quickly looking up countries using .loc method.

In [5]: df_can.set_index('Country', inplace=True)

# Let's view the first five elements and see how the dataframe was changed
df_can.head()

Out[5]: Continent Region DevName 1980 1981 1982 1983 1984 1985 1986

Country

Southern Developing
Afghanistan Asia 16 39 39 47 71 340 496
Asia regions

Southern Developed
Albania Europe 1 0 0 0 0 0 1
Europe regions

Northern Developing
Algeria Africa 80 67 71 69 63 44 69
Africa regions

American Developing
Oceania Polynesia 0 1 0 0 0 0 0
Samoa regions

Southern Developed
Andorra Europe 0 0 0 0 0 0 2
Europe regions

5 rows × 38 columns
 Notice now the country names now serve as indices.

In [6]: print('data dimensions:', df_can.shape)

data dimensions: (195, 38)

In [7]: # finally, let's create a list of years from 1980 - 2013

# this will come in handy when we start plotting the data
years = list(map(str, range(1980, 2014)))
#years = np.arange(1980,2014)

Line Plot
A line plot displays the relationship between two continuous variables over a continuous
interval, showing the trend or pattern of the data.

Let's created a line plot to visualize the immigrants (to Canada) trend during 1980 to 2013.
We need the Total of year-wise immigrants,

We will create a new dataframe for only columns containing the years
then, we will apply sum() on the dataframe

You can do create a line plot directly on axes by calling plot function plot()

In [8]: #As years is in the array format, you will be required to map it to str for plottin
#y=list(map(str, years))

#creating df with only years columns from 1980 - 2013

df_line=df_can[years]

#Applying sum to get total immigrants year-wise

total_immigrants=df_line.sum()
total_immigrants
Out[8]: 1980 99137
1981 110563
1982 104271
1983 75550
1984 73417
1985 69978
1986 86048
1987 134771
1988 139306
1989 164432
1990 188054
1991 207509
1992 221687
1993 222049
1994 193665
1995 187712
1996 200085
1997 192885
1998 160727
1999 179818
2000 216712
2001 242643
2002 223111
2003 217297
2004 232083
2005 257457
2006 247057
2007 232405
2008 243047
2009 248768
2010 276956
2011 246194
2012 256222
2013 257537
dtype: int64

In [14]: #Create figure and axes

fig, ax = plt.subplots()
plt.figure(figsize=(15, 6))
# Plot the line
ax.plot(total_immigrants)

#Setting up the Title

ax.set_title('Immigrants between 1980 to 2013')
#Setting up the Labels
ax.set_xlabel('Years')
ax.set_ylabel('Total Immigrants')

#Display the plot

plt.show()
 <Figure size 1500x600 with 0 Axes>

The plot function populated the x-axis with the index values (years), and
the y-axis with the column values (population).
However, notice how the years were not displayed because they are of type
string.

Therefore, let's change the type of the index values to integer for plotting.

In [10]: #Create figure and axes

fig, ax = plt.subplots()

#Changing the index type to integer

total_immigrants.index = total_immigrants.index.map(int)

# Plot the line

ax.plot(total_immigrants)

#Setting up the Title

ax.set_title('Immigrants between 1980 to 2013')

#Setting up the Labels

ax.set_xlabel('Years')
ax.set_ylabel('Total Immigrants')

#Display the plot

plt.show()
 Let's now customize the above plot's appearance

In [11]: #Create figure and axes

fig, ax = plt.subplots()

#Changing the index type to integer

total_immigrants.index = total_immigrants.index.map(int)

# Customizing the appearance of Plot

ax.plot(total_immigrants,
marker='s', #Including markers in squares shapes
markersize=5, #Setting the size of the marker
color='green', #Changing the color of the line
linestyle="dotted") #Changing the line style to a Dotted line
#Setting up the Title
ax.set_title('Immigrants between 1980 to 2013')

#Setting up the Labels

ax.set_xlabel('Years')
ax.set_ylabel('Total Immigrants')
ax.legend(['Immigrants'])

plt.show()
 Let's include the background grid, a legend and try to change the limits on
the axis

In [12]: #Create figure and axes

fig, ax = plt.subplots()

# Plot the line

ax.plot(total_immigrants,
marker='s', #Including markers in squares shapes
markersize=5, #Setting the size of the marker
color='green', #Changing the color of the line
linestyle="dotted") #Changing the line style to a Dotted line

#Setting up the Title

ax.set_title('Immigrants between 1980 to 2013')

#Setting up the Labels

ax.set_xlabel('Years')
ax.set_ylabel('Total Immigrants')

#limits on x-axis
plt.xlim(1975, 2015) #or ax.set_xlim()

#Enabling Grid
plt.grid(True) #or ax.grid()

#Legend
plt.legend(["Immigrants"]) #or ax.legend()
#Display the plot
plt.show()

Let's start with a case study:

In 2010, Haiti suffered a catastrophic magnitude 7.0 earthquake. The quake caused
widespread devastation and loss of life and aout three million people were affected by this
natural disaster. As part of Canada's humanitarian effort, the Government of Canada stepped
up its effort in accepting refugees from Haiti. We can quickly visualize this effort using a Line
plot:

Question: Plot a line graph of immigration from Haiti

You be required to create a dataframe where the name of the 'Country' is equal to 'Haiti'
and years from 1980 - 2013
Also you will be required to transpose the new dataframe in to a series for plotting
Might also have to change the type of index of the series to integer for a better look of the
plot
Then create fig and ax and call function plot() on the data.

Click here for a sample python solution

In [15]: #Creating data for plotting
df_can.reset_index(inplace=True)
haiti=df_can[df_can['Country']=='Haiti']

#creating haiti with only years columns from 1980 - 2013

#and transposing to get the result as a series
haiti=haiti[years].T

#converting the index to type integer

haiti.index = haiti.index.map(int)

#Plotting the line plot on the data

fig, ax = plt.subplots()
ax.plot(haiti)
#Setting up the Title
ax.set_title('Immigrants from Haiti between 1980 to 2013')
#Setting up the Labels
ax.set_xlabel('Years')
ax.set_ylabel('Number of Immigrants')
#Enabling Grid
#plt.grid(True) #or ax.grid()
#Legend
plt.legend(["Immigrants"]) #or ax.legend()
#Display the plot
plt.show()

We can clearly notice how number of immigrants from Haiti spiked up

from 2010 as Canada stepped up its efforts to accept refugees from Haiti.
 Let's annotate this spike in the plot by using the ax.annotate() method.

In [16]: fig, ax = plt.subplots()

ax.plot(haiti)

#Setting up the Title

ax.set_title('Immigrants from Haiti between 1980 to 2013')

#Setting up the Labels

ax.set_xlabel('Years')
ax.set_ylabel('Number of Immigrants')

#Enabling Grid and ticks

#plt.grid(True) #or ax.grid()
#ax.set_xticks(list(range(n, m, s)))

#Legend
plt.legend(["Immigrants"]) #or ax.legend()

ax.annotate('2010 Earthquake',xy=(2000, 6000))

plt.show()

You can also specify the ticks to be displayed on the plot like this -
ax.set_xticks(list(range(1980, 2015,5)))
 Scatter Plot
A scatter plot visualizes the relationship between two continuous variables, displaying
individual data points as dots on a two-dimensional plane, allowing for the examination of
patterns, clusters, and correlations.

Let's created a Scatter plot to visualize the immigrants (to Canada) trend during 1980 to
2013.
We need the Total of year-wise immigrants,

We will create a new dataframe for only columns containing the years
then, we will apply sum() on the dataframe

You can do create a scatter plot directly on ax by calling plot function scatter()

In [17]: #Create figure and axes

fig, ax = plt.subplots(figsize=(8, 4))

# Customizing Scatter Plot

ax.scatter(years, total_immigrants,
marker='o', #setting up the markers
s = 20, #setting up the size of the markers
color='darkblue')#the color for the marker

#add title
plt.title('Immigrants between 1980 to 2013')
#add labels
plt.xlabel('Years')
plt.ylabel('Total Immigrants')
#including grid
plt.grid(True)

#Legend at upper center of the figure

ax.legend(["Immigrants"], loc='upper center')

#Display the plot

plt.show()
In [18]: #Create figure and axes
fig, ax = plt.subplots(figsize=(8, 4))

total_immigrants.index = total_immigrants.index.map(int)

# Customizing Scatter Plot

ax.scatter(total_immigrants.index, total_immigrants,
marker='o', #setting up the markers
s = 20, #setting up the size of the markers
color='darkblue')#the color for the marker

#add title
plt.title('Immigrants between 1980 to 2013')
#add labels
plt.xlabel('Years')
plt.ylabel('Total Immigrants')
#including grid
plt.grid(True)

#Legend at upper center of the figure

ax.legend(["Immigrants"], loc='upper center')

#Display the plot

plt.show()
 Refer to the matplotlib documentation and change the
marker and its size, color in the above code to see the
difference in the appearance of the plot

Bar Plot
A bar plot represents categorical data with rectangular bars, where the height of each bar
corresponds to the value of a specific category, making it suitable for comparing values
across different categories.

Let's create a bar plot to visualize the top 5 countries that contribued the most immigrants
to Canada from 1980 to 2013.

Apply sort_values function on the 'Total' column of our data

We will create a new dataframe for only columns containing the years
then, we will apply sum() on the dataframe and can create a separatedataframe for top five
countries

You can further use the names of the countries to label each bar on the plot

In [19]: #Sorting the dataframe on 'Total' in descending order

df_can.sort_values(['Total'], ascending=False, axis=0, inplace=True)

# get the top 5 entries with head function

df_top5 = df_can.head()
 #resetting the index back to original way
df_bar_5=df_top5.reset_index()

#Creating alist of names of the top 5 countries

label=list(df_bar_5.Country)
label

Out[19]: ['India',
'China',
'United Kingdom of Great Britain and Northern Ireland',
'Philippines',
'Pakistan']

The third name is too lengthy to fit on the x-axis as label. Let's fix this using indexing

In [20]: label[2]='UK'
label

Out[20]: ['India', 'China', 'UK', 'Philippines', 'Pakistan']

In [21]: fig, ax = plt.subplots(figsize=(10, 4))

ax.bar(label,df_bar_5['Total'], label=label)
ax.set_title('Immigration Trend of Top 5 Countries')
ax.set_ylabel('Number of Immigrants')
ax.set_xlabel('Years')

plt.show()

Question:
Create a bar plot of the 5 countries that contributed the least to immigration to
Canada from 1980 to 2013.
In [22]: #Sorting the dataframe on 'Total' in descending order
df_can.sort_values(['Total'], ascending=True, axis=0, inplace=True)

# get the top 5 entries with head function

df_least5 = df_can.head()

#resetting the index back to original way

df_least5_bar=df_least5.reset_index()

#Creating alist of names of the top 5 countries

label=list(df_least5_bar.Country)
#label

fig, ax = plt.subplots(figsize=(10, 4))

ax.bar(label, df_least5_bar['Total'],label=label)
ax.set_title('Immigration Trend of Top 5 Countries')
ax.set_ylabel('Number of Immigrants')
ax.set_xlabel('Years')

plt.show()

Histogram
A histogram is a way of representing the frequency distribution of numeric dataset. The
way it works is it partitions the x-axis into bins, assigns each data point in our dataset to a
bin, and then counts the number of data points that have been assigned to each bin. So
the y-axis is the frequency or the number of data points in each bin. Note that we can
change the bin size and usually one needs to tweak it so that the distribution is displayed
nicely.

Let's find out the frequency distribution of the number (population) of new immigrants
from the various countries to Canada in 2013?
In [23]: df_country = df_can.groupby(['Country'])['2013'].sum().reset_index()

#Create figure and axes

fig, ax = plt.subplots(figsize=(10, 4))
ax.hist(df_country['2013'])
ax.set_title('New Immigrants in 2013')
ax.set_xlabel('Number of Immigrants')
ax.set_ylabel('Number of Countries')
ax.legend(['Immigrants'])

#Display the plot

plt.show()

Our plot doesnot match with the bars

By default, the histrogram method breaks up the dataset into 10 bins. The figure
below summarizes the bin ranges and the frequency distribution of immigration in 2013
The hist function retuns list of arrays with 1. counts and 2. bins. we can fetch that using
unpacking functionality
and further use the bins as x-ticks

In [24]: # Plot the bar

fig, ax = plt.subplots(figsize=(10, 4))
count = ax.hist(df_country['2013'])

#you can check the arrays in count with indexing count[0] for count, count[1] fo

ax.set_title('New Immigrants in 2013')

ax.set_xlabel('Number of Immigrants')
ax.set_ylabel('Number of Countries')
ax.set_xticks(list(map(int,count[1])))
ax.legend(['Immigrants'])

#Display the plot

plt.show()
 We can also plot multiple histograms on the same plot. For example,
let's try to answer the following questions using a histogram.

What is the immigration distribution for Denmark, Norway, and Sweden

for years 1980 - 2013?

In [25]: # let's quickly view the dataset

df=df_can.groupby(['Country'])[years].sum()
df_dns=df.loc[['Denmark', 'Norway', 'Sweden'], years]
df_dns=df_dns.T
df_dns
Out[25]: Country Denmark Norway Sweden

1980 272 116 281

1981 293 77 308

1982 299 106 222

1983 106 51 176

1984 93 31 128

1985 73 54 158

1986 93 56 187

1987 109 80 198

1988 129 73 171

1989 129 76 182

1990 118 83 130

1991 111 103 167

1992 158 74 179

1993 186 92 203

1994 93 60 192

1995 111 65 176

1996 70 70 161

1997 83 104 151

1998 63 31 123

1999 81 36 170

2000 93 56 138

2001 81 78 184

2002 70 74 149

2003 89 77 161

2004 89 73 129

2005 62 57 205

2006 101 53 139

2007 97 73 193

2008 108 66 165

2009 81 75 167
 Country Denmark Norway Sweden

2010 92 46 159

2011 93 49 134

2012 94 53 140

2013 81 59 140

In [26]: #Create figure and axes

fig, ax = plt.subplots(figsize=(10, 4))
ax.hist(df_dns)
ax.set_title('Immigration from Denmark, Norway, and Sweden from 1980 - 2013')
ax.set_xlabel('Number of Immigrants')
ax.set_ylabel('Number of Years')
ax.legend(['Denmark', 'Norway', 'Sweden'])
#Display the plot
plt.show()

Question:
What is the immigration distribution for China and India for years 2000 to 2013?

In [27]: # let's quickly view the dataset

df=df_can.groupby(['Country'])[years].sum()
y=list(map(str,range(2000, 2014)))
df_ci=df.loc[['China', 'India'], y]
df_ci=df_ci.T
#df_ci
fig, ax = plt.subplots(figsize=(10, 4))
ax.hist(df_ci)
ax.set_title('Immigration from Denmark, Norway, and Sweden from 1980 - 2013')
ax.set_xlabel('Number of Immigrants')
ax.set_ylabel('Number of Years')
 ax.legend(['China', 'India'])
#Display the plot
plt.show()

Click here for a sample python solution

Pie Chart
A pie chart represents the proportion or percentage distribution of different categories
in a dataset using sectors of a circular pie.
Let's create a pie chart representing the 'Total Immigrants' for the year 1980 to 1985

In [28]: fig,ax=plt.subplots()

#Pie on immigrants
ax.pie(total_immigrants[0:5], labels=years[0:5],
colors = ['gold','blue','lightgreen','coral','cyan'],
autopct='%1.1f%%',explode = [0,0,0,0,0.1]) #using explode to highlight t

ax.set_aspect('equal') # Ensure pie is drawn as a circle

plt.title('Distribution of Immigrants from 1980 to 1985')

#plt.legend(years[0:5]), include legend, if you donot want to pass the labels
plt.show()
 Question:
Create a pie chart representing the total immigrants proportion for each continent

First, you will have to group the data over continents and get the sum on total. Then
you can pass this data to the pie function

In [29]: #Creating data for plotting pie

df_con=df_can.groupby('Continent')['Total'].sum().reset_index()
label=list(df_con.Continent)
label[3] = 'LAC'
label[4] = 'NA'
df_con

Out[29]: Continent Total

0 Africa 618948

1 Asia 3317794

2 Europe 1410947

3 Latin America and the Caribbean 765148

4 Northern America 241142

5 Oceania 55174
In [30]: fig,ax=plt.subplots(figsize=(10, 4))

#Pie on immigrants
ax.pie(df_con['Total'], colors = ['gold','blue','lightgreen','coral','cyan','r
autopct='%1.1f%%', pctdistance=1.25)

ax.set_aspect('equal') # Ensure pie is drawn as a circle

plt.title('Continent-wise distribution of immigrants')

ax.legend(label,bbox_to_anchor=(1, 0, 0.5, 1))
plt.show()

Sub-plotting
Let us explore how to display more than one plot on the same figure and specify the
number of rows and columns to be created to the subplots function.

For instance, let’s create a line and scatter plot in one row
plt.subplots()
You can use the same functions using which you plotte lne and scatter plots at the
start of this lab

Both the subplots will be sharing the same y-axis as the data in the y-axis is the same.
So, assign the ‘Sharey’ parameter as True in the code below. Also notice the use of
'suptitle'
In [31]: # Create a figure with two axes in a row

fig, axs = plt.subplots(1, 2, sharey=True)

#Plotting in first axes - the left one

axs[0].plot(total_immigrants)
axs[0].set_title("Line plot on immigrants")

#Plotting in second axes - the right one

axs[1].scatter(total_immigrants.index, total_immigrants)
axs[1].set_title("Scatter plot on immigrants")

axs[0].set_ylabel("Number of Immigrants")

#Adding a Title for the Overall Figure

fig.suptitle('Subplotting Example', fontsize=15)

# Adjust spacing between subplots

fig.tight_layout()

# Show the figure

plt.show()

You can also implement the subplotting with add_subplot() as below:-

In [32]: # Create a figure with Four axes - two rows, two columns
fig = plt.figure(figsize=(8,4))
 # Add the first subplot (top-left)
axs1 = fig.add_subplot(1, 2, 1)
#Plotting in first axes - the left one
axs1.plot(total_immigrants)
axs1.set_title("Line plot on immigrants")

# Add the second subplot (top-right)

axs2 = fig.add_subplot(1, 2, 2)
#Plotting in second axes - the right one
axs2.barh(total_immigrants.index, total_immigrants) #Notice the use of 'barh'
axs2.set_title("Bar plot on immigrants")

#Adding a Title for the Overall Figure

fig.suptitle('Subplotting Example', fontsize=15)

# Adjust spacing between subplots

fig.tight_layout()

# Show the figure

plt.show()

Question: Choose any four plots, which you have

developed in this lab, with subplotting display them in a
2x2 display
In [33]: # Create a figure with Four axes - two rows, two columns
fig = plt.figure(figsize=(10, 10))

# Add the first subplot (top-left)

ax1 = fig.add_subplot(2, 2, 1)
ax1.plot(total_immigrants)
ax1.set_title('Plot 1 - Line Plot')

# Add the second subplot (top-right)

ax2 = fig.add_subplot(2, 2, 2)
ax2.scatter(total_immigrants.index, total_immigrants)
ax2.set_title('Plot 2 - Scatter plot')

# Add the third subplot (bottom-left)

ax3 = fig.add_subplot(2, 2, 3)
ax3.hist(df_dns)
ax3.set_title('Plot3 - Histogram')
ax3.set_xlabel('Number of Immigrants')
ax3.set_ylabel('Number of Years')

# Add the fourth subplot (bottom-right)

ax4 = fig.add_subplot(2, 2, 4)
ax4.pie(total_immigrants[0:5], labels=years[0:5],
colors = ['gold','blue','lightgreen','coral','cyan'],
autopct='%1.1f%%')
ax4.set_aspect('equal')
ax4.set_title('Plot 5 - Pie Chart')

#Adding a Title for the Overall Figure

fig.suptitle('Four Plots in a Figure Example', fontsize=15)

# Adjust spacing between subplots

fig.tight_layout()

# Show the figure

plt.show()
Congratulations! you have completed this lab!

Author
Dev Agnihotri