Data Visualization: Unlocking insights of Data

Dr. Aammar Tufail
Data Visualization

In today’s data-driven world, we are overwhelmed with vast amounts of information. To make sense of this deluge, Data Visualization has emerged as a critical tool. It transforms complex datasets into intuitive visuals, enabling better comprehension, decision-making, and insights. Dive in with us as we explore the power and potential of visualizing data!

In this post,

  • First we will learn what is Data Visualization and why do we need to be an expert in data visualization in order to jump into the field of Data Science/Analytic?
  • Secondly, we will learn about popular data visualization tools
  • Thirdly, we will see about the types of plots
  • Lastly, we will learn how to make those plots on dummy dataset using python libraries (pandas, matplotlib, seaborn, plotly, etc.).

Let’s have a look into it:

What is Data Visualization?

Data Visualization is the representation of data in a graphical format. It allows users to see and understand patterns, trends, and insights in data. From simple bar charts to intricate 3D models, data visualization encompasses a wide range of techniques.

Why is Data Visualization Important?

  1. Quick Insights: A well-crafted visual can communicate complex data in seconds, saving time and effort.
  2. Informed Decision Making: Visual data aids in making evidence-based decisions, reducing risks.
  3. Engagement: Visuals are more engaging than spreadsheets, making it easier to present and share findings.
  4. Data Exploration: Visualization tools enable users to delve deep into data, uncovering hidden insights.

Key Principles of Effective Data Visualization:

  1. Simplicity: Less is more. Avoid clutter and focus on the essentials.
  2. Consistency: Use consistent colors, fonts, and symbols to avoid confusion.
  3. Accuracy: Ensure the visual accurately represents the data.
  4. Interactivity: Modern tools allow for interactive visuals, enhancing user engagement and understanding.

Popular Data Visualization Tools:

  1. Tableau: A powerful tool offering a wide range of visualization options.
  2. Power BI: Microsoft’s solution for data visualization and business intelligence.
  3. D3.js: A JavaScript library for creating custom, dynamic visuals.
  4. Python Libraries: Libraries like Matplotlib, Seaborn, and Plotly are popular among data scientists.
  5. R Programming: ggplot2, plotly

Data Visualization in Action:

Imagine you’re a retailer looking to understand sales trends. A spreadsheet with thousands of rows of sales data might be overwhelming. However, a heat map showing sales density or a line chart depicting monthly sales trends can provide clear insights in moments.

For our readers, here’s a glimpse of the kind of visuals we’re talking about:

Different Plots-1
Geographical Maps
Bar plots
Dashboard
Previous slide
Next slide

The images above perfectly encapsulate the essence of data visualization. From analyzing pie charts to understanding sales trends through heat maps and line charts, these visuals demonstrate the power and versatility of data visualization tools. Moreover, the intricacy of a comprehensive dashboard showcases how various visualizations can come together to provide a holistic view of data.

Incorporating such visuals into your content not only enhances the reader’s experience but also solidifies their understanding of the topic. As the saying goes, “A picture is worth a thousand words,” and in the realm of data, a well-crafted visualization might be worth a million data points!

Harness the potential of data visualization, and let your data come alive. Happy visualizing! 🌟📊🖼️

Key Terms used in Data Visualization

Here’s a list of key terms in data visualization, along with explanations and examples for each:

  1. Bar Chart:

    • Explanation: A graphical representation of data using bars of varying heights or lengths.
    • Example: Comparing the sales of different products in a month.

  2. Pie Chart:

    • Explanation: A circular chart divided into slices to illustrate numerical proportions.
    • Example: Showing the market share of various smartphone brands.

  3. Histogram:

    • Explanation: A representation of the distribution of a dataset, similar to a bar chart but for frequency distribution.
    • Example: Displaying the age distribution of employees in a company.

  4. Scatter Plot:

    • Explanation: A graph with points plotted to show the relationship between two sets of data.
    • Example: Comparing advertising spend with sales revenue.

  5. Line Chart:

    • Explanation: A chart that displays information as a series of data points connected by straight line segments.
    • Example: Tracking stock market prices over a week.

  6. Heat Map:

    • Explanation: A data visualization technique where values in a matrix are represented as colors.
    • Example: Showing website activity, where darker colors represent more clicks.

  7. Box Plot (or Whisker Plot):

    • Explanation: A standardized way of displaying the dataset based on a five-number summary: minimum, first quartile, median, third quartile, and maximum.
    • Example: Comparing exam scores of different classes.

  8. Area Chart:

    • Explanation: Similar to a line chart, but the area between the axis and the line is filled with color or shading.
    • Example: Displaying the total revenue of a company over several years.

  9. Radar (or Spider) Chart:

    • Explanation: A chart that displays multivariate data on axes starting from the same point.
    • Example: Comparing the features of different products.

  10. Treemap:

  • Explanation: A visualization of hierarchical data using nested rectangles.
  • Example: Showing storage used by different types of files on a computer.
  1. Geographical Map:
  • Explanation: A map that displays data based on geographical areas or locations.
  • Example: Highlighting areas with high crime rates in a city.
  1. Time Series:
  • Explanation: A sequence of data points indexed in time order.
  • Example: Analyzing the daily temperature of a place over a year.

There are several other plots, such as bubble chart, violin plots and much more.

Plotting in Python

We can plot in python using libraries such as pandas, matplotlib, seaborn, plotly, altair, bokeh

1. Bar Chart

				
					# import pandas and matplotlib.pyplot libraries
import pandas as pd
import matplotlib.pyplot as plt

# create a dictionary with sample data
data = {'Products': ['Product A', 'Product B', 'Product C'],
    'Sales': [100, 150, 80]}

# create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# plot the dataframe as a bar chart
df.plot(x='Products', y='Sales', kind='bar', legend=False)

# set the y-axis label
plt.ylabel('Sales')

# set the title of the plot
plt.title('Sales of different products')

# save the plot to a file
plt.savefig('bar_chart.png', dpi=300, bbox_inches='tight')

# display the plot
plt.show()
bar_chart

2. Pie Chart

				
					# import pandas and matplotlib.pyplot libraries
import pandas as pd
import matplotlib.pyplot as plt

# create a dictionary with sample data
data = {'Products': ['Product A', 'Product B', 'Product C'],
        'Sales': [100, 150, 80]}

# create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# plot the dataframe as a pie chart
df.set_index('Products')['Sales'].plot(kind='pie', autopct='%1.1f%%', startangle=90)

# set the title of the plot
plt.title('Sales distribution of different products')

# save the plot to a file
plt.savefig('pie_chart.png', dpi=300, bbox_inches='tight')

# display the plot
plt.ylabel('')  # This is to remove the 'Sales' ylabel which is unnecessary in a pie chart
plt.show()
pie_chart

3. Histogram

				
					# import pandas and matplotlib.pyplot libraries
import pandas as pd
import matplotlib.pyplot as plt

# create a dictionary with sample data
data = {'Sales': [30,50,120,150,200,300,100, 150, 80, 120, 140, 130, 
                  50,50,30,20,10,50,35,45,55,50,55,65,56,58,58,54,55,50,5,
                  110, 145, 105, 90, 115, 125, 135, 85, 95]}

# create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# plot the dataframe as a histogram
df['Sales'].plot(kind='hist', bins=5, edgecolor='black')

# set the x-axis and y-axis labels
plt.xlabel('Sales Value')
plt.ylabel('Number of Transactions')

# set the title of the plot
plt.title('Distribution of Sales Transactions')

# save the plot to a file
plt.savefig('histogram.png', dpi=300, bbox_inches='tight')

# display the plot
plt.show()
histogram

4. Scatter plot

				
					# import pandas and matplotlib.pyplot libraries
import pandas as pd
import matplotlib.pyplot as plt

# create a dictionary with sample data
data = {
    'Sales': [100, 150, 80, 120, 140, 130, 110, 145, 105, 90],
    'Profit': [50, 70, 40, 60, 75, 65, 55, 72, 52, 45]
}

# create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# plot the dataframe as a scatter plot
df.plot(x='Sales', y='Profit', kind='scatter')

# set the x-axis and y-axis labels
plt.xlabel('Sales Value')
plt.ylabel('Profit Value')

# set the title of the plot
plt.title('Relationship between Sales and Profit')

# save the plot to a file
plt.savefig('scatter_plot.png', dpi=300, bbox_inches='tight')

# display the plot
plt.show()
scatter_plot

5. Line Chart

				
					# import pandas and matplotlib.pyplot libraries
import pandas as pd
import matplotlib.pyplot as plt

# create a dictionary with sample data
data = {
    'Month': ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'],
    'Monthly Sales': [100, 110, 105, 115, 120, 125, 130, 135, 140, 145, 150, 155]
}

# create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# plot the dataframe as a line chart
df.plot(x='Month', y='Monthly Sales', kind='line', marker='o')

# set the x-axis and y-axis labels
plt.xlabel('Month')
plt.ylabel('Sales Value')

# set the title of the plot
plt.title('Monthly Sales Trend over a Year')

# save the plot to a file
plt.savefig('line_chart.png', dpi=300, bbox_inches='tight')

# display the plot
plt.show()
line_chart

6. Heatmap

				
					# import required libraries
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Sample correlation matrix
data = {
    'Product A': [1, 0.6, 0.8, 0.3],
    'Product B': [0.6, 1, 0.5, 0.7],
    'Product C': [0.8, 0.5, 1, 0.4],
    'Product D': [0.3, 0.7, 0.4, 1]
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Plotting the heatmap
plt.figure(figsize=(8, 6))
sns.heatmap(df, annot=True, cmap='viridis', vmin=0, vmax=1)

# Set the title of the plot
plt.title('Product Correlation Heatmap')

# Save the plot to a file
plt.savefig('heatmap.png', dpi=300, bbox_inches='tight')

# Display the plot
plt.show()
heatmap

7. Boxplot

				
					# import required libraries
import pandas as pd
import matplotlib.pyplot as plt

# Sample data
data = {
    'Product A': [95, 85, 90, 100, 105, 110],
    'Product B': [100, 105, 95, 90, 115, 120],
    'Product C': [105, 110, 115, 120, 125, 130]
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Plotting the boxplot
df.boxplot(grid=False, vert=True, fontsize=10)

# Set the title of the plot
plt.title('Sales Distribution by Product')

# Set the y-axis label
plt.ylabel('Sales Value')

# Save the plot to a file
plt.savefig('boxplot.png', dpi=300, bbox_inches='tight')

# Display the plot
plt.show()
boxplot

Box plot can also be created using seaborn library as follows

				
					# we can do the same thing with seaborn library as follows

# import required libraries
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Sample data
data = {
    'Product A': [95, 85, 90, 100, 105, 110],
    'Product B': [100, 105, 95, 90, 115, 120],
    'Product C': [105, 110, 115, 120, 125, 130]
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Plotting the boxplot using seaborn
plt.figure(figsize=(10, 6))
sns.boxplot(data=df)

# Set the title of the plot
plt.title('Sales Distribution by Product')

# Set the y-axis label
plt.ylabel('Sales Value')

# Save the plot to a file
plt.savefig('sns_boxplot.png', dpi=300, bbox_inches='tight')

# Display the plot
plt.show()
sns_boxplot

8. Area Chart

				
					# import required libraries
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Sample data
data = {
    'Month': ['Jan', 'Feb', 'Mar', 'Apr', 'May'],
    'Sales': [100, 150, 180, 220, 270]  # Accumulated sales for demonstration
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Plotting the area chart
plt.figure(figsize=(10, 6))
sns.set_style("whitegrid")
plt.fill_between(df['Month'], df['Sales'], color="skyblue", alpha=0.4)
plt.plot(df['Month'], df['Sales'], color="Slateblue", alpha=0.6)

# Set the title, x-axis and y-axis labels
plt.title('Monthly Accumulated Sales')
plt.xlabel('Month')
plt.ylabel('Accumulated Sales Value')

# Save the plot to a file
plt.savefig('area_chart.png', dpi=300, bbox_inches='tight')

# Display the plot
plt.show()
area_chart

9. Spider or Radar Plot

				
					# import required libraries
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

# Sample data
data = {
    'Metrics': ['Metric A', 'Metric B', 'Metric C', 'Metric D', 'Metric E'],
    'Product 1': [3, 4, 5, 4, 2],
    'Product 2': [4, 5, 4, 3, 3]
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Number of variables
categories = list(df['Metrics'])
N = len(categories)

# Set the angle for each metric's axis
angles = [n / float(N) * 2 * np.pi for n in range(N)]
angles += angles[:1]

# Initialize the spider plot
plt.figure(figsize=(8, 6))
ax = plt.subplot(111, polar=True)

# Plotting for Product 1
values = df['Product 1'].tolist()
values += values[:1]  # Repeat the first value to close the circular graph
ax.plot(angles, values, linewidth=2, linestyle='solid', label='Product 1')
ax.fill(angles, values, 'b', alpha=0.1)

# Plotting for Product 2
values = df['Product 2'].tolist()
values += values[:1]  # Repeat the first value to close the circular graph
ax.plot(angles, values, linewidth=2, linestyle='solid', label='Product 2')
ax.fill(angles, values, 'r', alpha=0.1)

# Add the axis labels
plt.xticks(angles[:-1], categories, color='grey', size=12)
ax.set_rlabel_position(30)
plt.yticks([1,2,3,4,5], ["1","2","3","4","5"], color="grey", size=8)
plt.ylim(0, 5)

# Add title and legend
plt.title('Comparison between Product 1 and Product 2 on various metrics')
ax.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))

# Save the plot to a file
plt.savefig('radar_chart.png', dpi=300, bbox_inches='tight')

# Display the plot
plt.show()
radar_chart

10. Tree Map

				
					# Import required libraries
import pandas as pd
import matplotlib.pyplot as plt
import squarify # install if you do not have it

# Sample data
data = {
    'File Type': ['Documents', 'Videos', 'Images', 'Audio', 'Other'],
    'Storage Used (GB)': [50, 150, 80, 60, 10]
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Plotting the treemap
plt.figure(figsize=(12, 8))
colors = ['red', 'green', 'blue', 'yellow', 'grey']
squarify.plot(sizes=df['Storage Used (GB)'], label=df['File Type'], color=colors, alpha=0.6)

# Add title
plt.title('Storage Used by File Type')
plt.axis('off')  # Turn off the axis

# Save the plot to a file
plt.savefig('treemap.png', dpi=300, bbox_inches='tight')

# Display the plot
plt.show()
treemap

11. Geographical Maps

				
					# Import required libraries
import plotly.graph_objects as go

# Data for cities and their coordinates
cities = {
    'New York': [40.7128, -74.0060],
    'London': [51.5074, -0.1278],
    'Sydney': [-33.8688, 151.2093],
    'Tokyo': [35.6895, 139.6917],
    'Rio de Janeiro': [-22.9068, -43.1729]
}

# Extracting latitudes, longitudes, and city names for plotting
lats = [coords[0] for coords in cities.values()]
longs = [coords[1] for coords in cities.values()]
names = list(cities.keys())

# Creating the Scattergeo plot
fig = go.Figure(data=go.Scattergeo(
    lon = longs,
    lat = lats,
    text = names,
    mode = 'markers',
    marker = dict(
        size = 10,
        opacity = 0.8,
        reversescale = True,
        autocolorscale = False,
        symbol = 'circle',
        line = dict(
            width=1,
            color='rgba(102, 102, 102)'
        ),
    )
))

# Setting the layout for the map
fig.update_layout(
    title = 'Cities on World Map using Plotly',
    geo = dict(
        scope='world',
        showland = True,
    )
)

# Display the map
fig.show()

# Save the map to an HTML file
fig.write_html("world_map_plotly.html")
geomap

12. Time Series plots

				
					# Import required libraries
import plotly.graph_objects as go
import pandas as pd

# Sample time series data
data = {
    'Date': ['2022-01-01', '2022-01-02', '2022-01-03', '2022-01-04', '2022-01-05'],
    'Value': [10, 15, 7, 12, 10]
}

# Create a pandas dataframe from the dictionary
df = pd.DataFrame(data)

# Convert the 'Date' column to a datetime object for better handling
df['Date'] = pd.to_datetime(df['Date'])

# Creating the time series plot
fig = go.Figure(data=go.Scatter(x=df['Date'], y=df['Value'], mode='lines+markers'))

# Setting the layout for the plot
fig.update_layout(
    title='Time Series Plot using Plotly',
    xaxis_title='Date',
    yaxis_title='Value',
    xaxis=dict(showline=True, showgrid=False, showticklabels=True),
    yaxis=dict(zeroline=False, showgrid=True, showline=False, showticklabels=True),
)

# Display the plot
fig.show()

# Save the plot to an HTML file
fig.write_html("time_series_plotly.html")

# save as png
fig.write_image("time_series_plotly.png")
time_series_plotly

13. Bubble Chart

				
					# Import required libraries
import plotly.graph_objects as go

# Sample data
data = {
    'X': [10, 20, 30, 40, 50],
    'Y': [15, 25, 35, 45, 55],
    'Size': [30, 60, 90, 120, 150],  # Determines the size of bubbles
    'Labels': ['A', 'B', 'C', 'D', 'E']
}

# Creating the bubble chart
fig = go.Figure(data=go.Scatter(
    x=data['X'],
    y=data['Y'],
    mode='markers',
    text=data['Labels'],
    marker=dict(
        size=data['Size'],
        sizemode='diameter',  # 'diameter' ensures that the size values correspond to the diameter of the bubbles
        opacity=0.6,
        color=data['Y'],  # Coloring the bubbles based on Y values
        colorscale='Viridis',
        colorbar=dict(title='Colorbar')
    )
))

# Setting the layout for the plot
fig.update_layout(
    title='Bubble Chart using Plotly',
    xaxis_title='X Values',
    yaxis_title='Y Values',
    showlegend=False
)

# Display the plot
fig.show()

# Save the plot to an HTML file
fig.write_html("bubble_chart_plotly.html")

# save png file with 300 dpi
fig.write_image("bubble_chart_plotly.png", scale=3)
bubble_chart_plotly

14. Violin plots

A violin plot is similar to a box plot but also includes a rotated kernel density plot on each side. It provides a visualization of the distribution of the data, its probability density, and its cumulative distribution.

Here’s how to create a violin plot using Plotly:

				
					# Import required libraries
import plotly.express as px

# Sample data
data = {
    'Category': ['A', 'B', 'C', 'A', 'B', 'C', 'A', 'B', 'C', 'A', 'B', 'C'],
    'Values': [10, 15, 5, 25, 20, 10, 30, 25, 5, 20, 15, 10]
}

# Creating the violin plot using plotly express
fig = px.violin(data, y="Values", x="Category", box=True, points="all")
# if box=False then no Boxplots will be there inside violins

# Setting the layout for the plot
fig.update_layout(
    title='Violin Plot using Plotly',
    yaxis_title='Values',
    xaxis_title='Category'
)

# Display the plot
fig.show()

# Save the plot to an HTML file
fig.write_html("violin_plot_plotly.html")

#save png file with 300 dpi
fig.write_image("violin_plot_plotly.png", scale=3)
violin_plot_plotly

Data Visualization tutorial in Urdu/Hindi

Learn from our free courses here:

Data Analyst Summer Bootcamp 2024
4.50
(6)

Data Analyst Summer Bootcamp 2024

Original price was: ₨30,000.00.Current price is: ₨22,500.00.
Six months of AI and Data Science Mentorship Program
4.94
(164)

Six months of AI and Data Science Mentorship Program

Python ka Chilla for Data Science (40 Days of Python for Data Science)
4.84
(62)

Python ka Chilla for Data Science (40 Days of Python for Data Science)

Start Learning

Our recent blogs

Websites must try

Powerful Websites

May 26, 2024 1 Comment

Advancements in website tools 💻 in the digital age provide invaluable resources. From e-commerce giants and social media

Read More »
google_scholar_for_advance_use

Google Scholar Kaise Use Karein?

May 4, 2024 No Comments

Google Scholar ek powerful search engine hai jo scientific literature ke liye use kiya jata hai. Is article mein hum aapko bataein ge ke Google Scholar ko kaise use karein aur kab kab aapko dusre tools ya databases ko use karna chahiye.

Read More »

Chapter 8: The Future of Sampling in Statistics – Emerging Trends and Innovations

November 30, 2023 5 Comments

The future of sampling in statistics is vibrant and full of potential, marked by technological innovation and methodological advancements. As we embrace these changes, sampling will continue to be a pivotal tool in unraveling the complexities of the world through data. The journey ahead is not just about statistical techniques; it’s about shaping a future where data is collected, analyzed, and used responsibly and innovatively.

Read More »

Let us know in your comment section which other plots would you like to make?

26 Comments.

  2. dear Ammar sb
    thanks for your efforts, i did not enroll your course but whenever i have free time i do watch your videos on youtube, after python ka chilla i started 6 month course of data science, i read this block and run the code given one by one all code runs perfectly excep last two that is time series and violin plots. it did not show any error just keep the code running but did not gave be result.

    Reply

  10. AOA,
    Data Visualization: Unlocking Insights of Data is very helpful to understand data visualization. I will practice all of them insha ALLAH.
    ALLAH KAREEM aap ko dono jahan ki bhaliya aata kry.AAMEEN

    Reply

  12. I need to work hard to find the difference between Plotting the Data Frames in every type of Plts, as the formulae seem to be the same for all but the Elements inside it are the main factors that decide which type of Plot this is going to make, that’s my key take away from this blog. Thanks Ammar!
    Best Regards,

    Reply

  17. Making great efforts from grass root to advanced levels with detailed coding and practical approach, salute sir. Ma Sha Allah, Best wishes ever.

    Reply

  26. This blog is an absolute treasure for data enthusiasts! It skillfully breaks down the world of data visualization using Python libraries like pandas, plotly, seaborn, and matplotlib. The clear explanations and practical code examples make it an invaluable resource for anyone looking to master the art of data visualization. Kudos to the author for simplifying complex concepts and making data visualization accessible to all!

    Reply

  27. Just wow! Boht sy students ko visualization k liye plots choose krne mein mushkil hoti lykin kamal kr diya👌🏻

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *