Data Science by Nireekshan

2. Data Visualization with matplotlib

matplotlib

 matplotlib is the most popular python plotting library.

 It has a module called pyplot that makes plotting easy by providing features for controlling
font properties, line styles, formatting axes, etc.
 Matplotlib is very good for creating graphs like line charts, bar charts, histograms and
many more.

1|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

1. Line charts introduction

 A line chart or line graph is a type of chart which displays information as a series of data
points connected by straight line
 A line chart is often used to visualize a trend in data over intervals of time.

Technical info

 plt.plot() - functions create a line plot

 plot.xlabel() & plot.ylabel() - helps to label x and y axis
 plot.legend() - helps to signify the observation variables.
 plot.title() - helps to set the title of the plot.
 plot.show() - helps to display the plot

Program create a simple line chart

Name demo1.py

import numpy as np
import matplotlib.pyplot as plt

x = np.arange(0, 10)
y=x+1

plt.plot(x,y)
plt.show()

Output

2|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

Program create a simple line chart

Name demo2.py

import numpy as np
import matplotlib.pyplot as plt

x = np.arange(0, 10)
y=x+2
z=x+3

plt.title("A Graph")

plt.plot(x, y)
plt.plot(x, z)
plt.show()

Output

3|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

Giving title to line

 We can label the axes as well and add title

Program create a simple line chart and giving title

Name demo3.py

import numpy as np
import matplotlib.pyplot as plt

x = np.arange(0, 10)
y=x+2

plt.title("A Graph")

plt.plot(x, y)
plt.show()

Output

4|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

Labelling the axes

 We can label x axis and y axis by using xlabel and ylabel

Program create a simple line chart and giving title and labelling
Name demo4.py

import numpy as np
import matplotlib.pyplot as plt

x = np.arange(0,10)
y=x+2

plt.title("A Graph")

plt.xlabel("X values")
plt.ylabel("Y values")

plt.plot(x,y)
plt.show()

Output

5|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

2. Multiple Plots

 With matplotlib, you can create more than one plot on the same canvas.
 You do so by use of the subplot() function which defines the location and the number of
the plot.

Program create a multiple plots

Name demo5.py

import matplotlib.pyplot as plt

x = range(0, 20)
y = range(0, 40, 2)

plt.subplot(2, 1, 1)

plt.plot(x, y)

plt.ylabel('Value')
plt.title('First chart')
plt.grid(True)

plt.subplot(2, 1, 2)

plt.plot(x, y)

plt.xlabel('Item (s)')
plt.ylabel('Value')
plt.title('Second chart')
plt.grid(True)

plt.show()

Output

 The subplot() command specifies numrows, numcols and fignum.

6|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

3. Stack Plot

 This is an advanced line chart or bar chart that breaks down data from various categories
and stacks them together so that a comparison between the values from various
categories may be made.
 Suppose you need to compare the sales scored by three different months per year over
the last 8 years.

Program Creating stack plot

Name demo6.py

import matplotlib.pyplot as plt

year = [2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018]

sales1_mobiles = [800,1000,1700,1500,2300,1800,2400,2900]
sales2_tvs = [1000,1400,1900,1600,2500,2000,2600,3200]
sales3_freeze = [1200,1700,2100,1900,2600,2200,2800,3500]

plt.plot([],[], color='y', label = ' sales1_mobiles ')

plt.plot([],[], color='r', label = ' sales2_tvs')
plt.plot([],[], color='b', label = 'sales3_freeze')

plt.stackplot(year, sales1_mobiles, sales2_tvs, sales3_freeze, colors = ['y', 'r', 'b'])

plt.legend()

plt.title('Sales Information')
plt.xlabel('year')
plt.ylabel('sales')

plt.show()

Output

7|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

4. Pie Chart

 This is a circular plot that has been divided into slices displaying numerical proportions.
 Every slice in the pie chart shows the proportion of the element to the whole.
 A large category means that it will occupy a larger portion of the pie chart.

Program Creating pie chart

Name demo7.py

import matplotlib.pyplot as plt

students = 'Nireekshan', 'Abhi', 'Srinu'

points = [62,48,36]

plt.pie(points, labels = students)

plt.axis('equal')
plt.show()

Output

8|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

Attributes

 To create a pie chart, we call the pie function of the pyplot module.
 The first parameter to the function is the list of numbers for every category.
o labels attribute:
 A list of categories separated by commas is then passed as the argument
to labels attribute.
o colors attribute:
 To provide the color for every category.
o To create shadows around the various categories in pie chart.
o To split each slice of the pie chart into its own.

Program Creating pie chart

Name demo8.py

import matplotlib.pyplot as plt

students = 'Nireekshan', 'Abhi', 'Srinu'

points = [60, 48, 36]
colors = ['y', 'r', 'b']

plt.pie(points, labels = students, colors=colors , shadow = True, explode = (0.05,

0.05, 0.05), autopct = '%1.1f%%')

plt.axis('equal')
plt.show()

Output

9|Page nireekshan.ds@gmail.com
 Data Science by Nireekshan

5. Histogram

 A histogram is an accurate graphical representation of the distribution of numerical data.

 It takes as input one numerical variable only.
 The variable is cut into several bins, and the number of observation per bin is represented
by the height of the bar.
 It is a good tool when you need to understand the count of data ranges.

Program Creating histogram

Name demo9.py

import pandas as pd
import matplotlib.pyplot as plt

titanic_data = pd.read_csv("sales8.csv")
titanic_data['Quantity'].hist()

plt.show()

Output

10 | P a g e nireekshan.ds@gmail.com
 Data Science by Nireekshan

6. Scatter Plot

 This is a type of plot that shows many data points plotted.

 Each point is a representation of two variables.
 One of the variables is chosen on the vertical axis while the other one is chosen on the
horizontal axis.
 To create a scatter plot, we call the scatter() method of pyplot module.
 This method takes two numeric data points for scattering the data points on the plot.

Program Creating Scatter plot

Name demo10.py

import pandas as pd
import matplotlib.pyplot as plt

titanic_data = pd.read_csv("titanic_train.csv")
titanic_data.plot.scatter(x='Age', y='Fare', figsize=(8,6))

plt.show()

Output

11 | P a g e nireekshan.ds@gmail.com
 Data Science by Nireekshan

7. Box Plots

 Box plots help us measure how well data in a dataset is distributed.

 The dataset is divided into three quartiles.
 The graph shows the maximum, minimum, median, first quartile and third quartiles of the
dataset.
 It is also good for comparing how data is distributed across datasets by creating box plots
for each dataset.

Use Box plots

 Use a boxplot when you need to get the overall statistical information about the data
distribution.
 It is a good tool for detecting outliers in a dataset.
 To create a boxplot, we call the boxplot function of pyplot.
 The function takes the name of the dataset as the parameter.

Program Creating box plot

Name demo11.py

import matplotlib.pyplot as plt

student1 = [8,10,17,15,23,18,24,29]
student2 = [10,14,19,16,25,20,26,32]
student3 = [12,17,21,19,26,22,28,35]

data=[student1, student2, student3]

plt.boxplot(data)

plt.show()

Output

 The line dividing the box into two shows the median of the data.
 The end of the box represents the upper quartile (75%) while the start of the box
represents the lower quartile (25%).
 The part between the upper quartile and the lower quartile is known as the Inter Quartile
Range (IQR) and helps in approximating 50% of the middle data.

12 | P a g e nireekshan.ds@gmail.com
 Data Science by Nireekshan

8. Bar Chart

 This type of chart is used for showing the distribution of data over many groups.
 Most people confuse it with the histogram but note that a histogram only accepts
numerical data for plotting.

When to use?

 It is only good for comparing numerical values.

 Use a bar plot when you need to make a comparison between multiple groups.
 To create a bar plot, we use the bar function of the Matplotlib library.

Program Creating bar chart

Name demo12.py

import matplotlib.pyplot as plt

months = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]

sales = [23, 45, 56, 78, 213, 45, 78, 89, 99, 100, 101, 130]

plt.bar(months, sales)

plt.xlabel('Month')
plt.ylabel('Product Sales')

plt.title('A Bar Graph')

plt.show()

Output

13 | P a g e nireekshan.ds@gmail.com
 Data Science by Nireekshan

9. Bubble Chart

 This type of chart shows the data in the form of a cluster of circles.
 The data to generate the bubble chart should have the xy coordinates, the bubble size and
the color of the bubbles.
 The colors can be supplied by use of the Matplotlib library.
 To create a bubble chart, we use the scatter function provided in the Matplotlib library.
Here is an example:

Program Creating bubble chart

Name demo13.py

import numpy as np
import matplotlib.pyplot as plt

x = np.random.rand(30)
y = np.random.rand(30)
z = np.random.rand(30)

colors = np.random.rand(30)

plt.scatter(x, y, s=z*1000,c=colors)

plt.show()

Output

14 | P a g e nireekshan.ds@gmail.com
 Data Science by Nireekshan

10. Heat Maps

 A heat map has values representing different shades of a similar color for every value that
is to be plotted.
 The darker shades of the chart indicate the higher values compared to the lighter shades.
 For the case of a very different value, you can use a more different color.

Program Creating heat maps

Name demo14.py

from pandas import DataFrame

import matplotlib.pyplot as plt

data=[{3, 4, 6, 1}, {6, 5, 4, 2}, {7, 3, 5, 2}, {2, 7, 5, 3}, {1, 8, 1, 4}]
Index= ['I1', 'I2','I3','I4','I5']

Cols = ['Col1', 'Col2', 'Col3','Col4']

df = DataFrame(data, index=Index, columns=Cols)

plt.pcolor(df)

plt.show()

Output

 We have created a two-dimensional plot of values that are mapped to the columns and
indices of the chart.

15 | P a g e nireekshan.ds@gmail.com