data visualization with python assignment coursera

Richard Lung Wicaksono

5+ years of experience in programming and data management. I coded in languages such as Python, Java, C++, HTML/CSS, SQL, PHP, and JavaScript. Conducted internships at PT. HK Metals Utama Tbk as a Data Analyst intern and Zi.Care as a Quality Assurance intern. Currently a Data Conversion Engineer at Nelnet, Inc., creating and developing existing SQL scripts to improve work efficiency and automation. Outside programming, my hobbies include watching and playing soccer, basketball, and badminton. View My LinkedIn Profile

View My GitHub Profile

Data Visualization with Python - Final Assignment

This was the final assignment for Data Visualization with Python by IBM, a course included in the IBM Data Science Professional Certificate program by Coursera. Developed in Python.

Project Details

This was the final assignment for Data Visualization with Python by IBM on Coursera.

There are two parts to this final project:

Part 1 - Create visualizations using Matplotlib, Seaborn, & Folium

The objective of this part of the Final Assignment is to analyze the historical trends in automobile sales during recession periods. The goal is to provide insights into how the sales of XYZAutomotives, a company specializing in automotive sales, were affected during times of recession.

Part 2 - Create Dashboard using Plotly and Dash

The objective of this part of the Fnal assignment is to create dashboards to contain your plots and charts and to provide the directors with the ability to select a particular report or a period of time so they can discuss the data in detail.

Project Outcome

There are two dropdowns that were created. The first dropdown lets the user select which report they wish to view: Recession Period Statistics or Yearly Statistics. The second dropdown lets the user select the year in which the data were recorded. If they chose the Recession Period Statistics report, however, they will not be able to select the year from the dropdown.

Charts and graphs for the Recession Period Statistics.

Charts and graphs for the Yearly Statistics report. Users are able to select a specific year from the second dropdown, and all the graphs and charts will be updated in real-time.

Here’s Your Guide to IBM’s “Data Visualization with Python” Final Assignment (or How to Code a Dashboard using Plotly Dash)

By Marvin Rubia on 2023-07-23

This article will guide you to accomplish the final assignment of Data Visualization with Python , a course created by IBM and offered by Coursera. Nevertheless, this tutorial is for anyone— enrolled in the course or not — who wants to learn how to code an interactive dashboard in Python using Plotly’s Dash library. I assume, however, that you already know how to use Pandas for analysis.

I want you to learn

If you just copy the codes in this article and ignore how to understand them, you miss the whole point of learning data visualization (or data science in general). Typing the codes on your own helps you be more aware of their syntax, logic, and structure. Also, learning by doing nudges you to ask relevant questions and search for their answers in your browser. This process pushes you to learn better.

What is the final assignment?

You will play a role of a data analyst tasked to monitor and report US domestic airline flights. Your goal is to analyze the performance of reporting airlines to improve flight reliability, thereby enhancing customer reliability. Your dashboard should have a dropdown for the type of report with two options:

• Yearly Airline Performance Report , and
• Yearly Airline Delay Report .

It should also have another dropdown that selects what Year (from 2005 to 2020) to report. The two pictures below (corresponding to the two report types) show my dashboard for 2009. The kinds of graphs for each report are instructed by the assignment.

💡 Speed up your blog creation with DifferAI . Available for free exclusively on the free and open blogging platform, Differ .

Where will I code?

I will code the Dash application in my local Jupyter Notebook. The course, however, invites the student to use IBM’s Skills Networks Labs Cloud IDE (which uses Theia Cloud IDE). Unfortunately, all of your work in this environment is lost when you close your session, timed out due to inactivity, or logged off.

Gladly, the final assignment is graded only on the screenshots of your dashboard (given a report type and a year), no matter how or where you coded it . Thus, I am going to use Jupyter Notebook, which works nicely with Dash 2.11 and later versions .

Note: My codes do not follow everything from the instructor’s code hints. You might also see that some of my graph titles and codes for layout customizations are different from the course’s expectations. What matters is that we can produce the desired dashboard. In the real world, most stakeholders do not care about the behind-the-scenes codes. What they expect is the output.

Basic components of a Dash app

• Layout = This is the structure of your dashboard. The codes for this prepare how your dashboard will look like.
• Callback Function = This is the function called by Dash whenever an input component’s property changes. This makes the dashboard interactive.

Necessary installations

Where does the data come from?

The dataset (in .csv format) is provided by IBM and it contains real data on US Domestic Flights from 1987 to 2020, taken from the US Bureau of Transportation Statistics. You can download the dataset from here . To know what each column means, refer to this link .

Let’s start coding!

Importing libraries

Preparing our data.

The output for this is:

Getting only 2005–2020 data (as per the instruction)

The output for the previous codes shows the first 5 rows of the 2005–2020 dataframe.

The airline_data is now our dataframe moving forward. We are going to plot 5 graphs for each type of report. In other words, we need to have a total of 10 subsets of our airline_data .

Function to get necessary data for performance report (Type 1)

Function to get necessary data for delay report (type 2).

Just like what I said in the introduction, I assume you already know how to manipulate dataframe using Pandas. Although, you can have a sense of each sub-dataframe by running them one by one.

Coding the dashboard’s layout

Note for desktop viewers: To see the remaining code on a given line, you can highlight any word and extend it to the right.

Read the subheading comments to understand the purpose of each chunk of codes above. Remember, we installed the dash-html-components , which allows us to produce HTML-like codes in Python. The html.Div() is a division in our dashboard. It tells Dash to allot a space for whatever we want to code inside that division. Below the heading html.H1 are 5 layers of divisions. When you run the previous codes, you’ll get something like this:

If you count the total html divisions from the layout code, we actually have a total of 10. But we were able to nest the two sub-divisions in the 2nd, 3rd, 5th, and 6th layers with the code style={‘display’: ‘flex’} .

In our layout, the inputs (report types and year) come from two dcc.Dropdown() components represented by their id parameters. Meanwhile, the id inside each dcc.Graph() component represents each graph, which is our output. The logic is this: the output graphs should respond to the change in input values as instructed by the callback function .

To learn more about Dash Dropdowns, refer to this documentation . To learn more about Dash Graphs, refer to this documentation .

Coding the dashboard’s callback decorator and function

We are going to add the callback function after our app.layout codes and before running the app . The following shows the callback decorator and the callback function.

The callback decorator tells Dash that we have 5 outputs from two inputs. Each of which has designated id we coded in app.layout() .

In contrast with the assignment’s instruction, I did not use State() objects in the callback operator. State() is similar to Input() , except that you have to finish updating all State() objects before the callback function is fired. (To read more about it, click here .) I believe it is not necessary for our dashboard. If I update the report type dropdown input, I am okay with seeing the graphs immediately change even if I have not updated yet the year dropdown input.

The first two lines of our callback function get_graphs(report_type, year) select the data from our airline_data dataframe with the given year input.

Our callback function returns either the list [tree_fig, pie_fig, map_fig, bar_fig, line_fig] if Yearly Airline Performance Report is chosen or the list [carrier_fig, weather_fig, nas_fig, sec_fig, late_fig] if Yearly Airline Delay Report is chosen. The order within either list matters . The order corresponds to the order of the Output(component_id) in the callback decorator.

The rest of the codes are the syntax for creating graphs using Plotly Express, which I assume you have learned from the course modules (or from Plotly documentations) prior to the final assignment.

Full Code (from creating the Dash app to running it)

You might notice that my last line above has a different syntax from the one taught by the instructor in running a Dash app. That’s because I coded it in Jupyter Notebook, and the app.run(jupyter_mode= ‘tab’) code automatically produces our dashboard in a new tab (instead of inline with the Notebook cell). This allows us to have a fullscreen view of our dashboard.

How will you be graded?

If you are enrolled in this course, you will be instructed to submit a screenshot of your dashboard for a given report type and year.

The primary purpose of this article is to help you learn how to code a dashboard in Python from its backend. The course certificate is just secondary. That said, thank you for reading this guide, and congratulations in advance on claiming your IBM certificate!

Update: I have now modified my dashboard by customizing the transparency of Plotly charts and adding a background image. It looks better. Check it out!

• Adding a Background Image to a Dash’s Dashboard

If you found value in this post, kindly click the clap button (many times?) and share this with your friends. You can also follow me on Medium.com for more articles on critical thinking, data analytics, machine learning, science and society, and more. Want to connect with me on LinkedIn? Find me here .

Continue Learning

Guide to python project structure and packaging, best python libraries for algorithmic trading.

In this article we help you to define which Python libraries work best if you are actively engaged with algorithmic trading with Python

Build a Complete Invoicing Web Application with Django

How to build a complete invoicing web application with Django — Final Tutorial

How to Make a Beautiful Donut Chart and Nested Donut Chart in Matplotlib

Donut charts are used to show the proportions of categorical data, with the size of each piece representing the proportion of each category.

Python 3.12: A Game-Changer in Performance and Efficiency

Exploring the Exciting New Additions, Updates, and Changes in Python 3.12

Analytical Differentiation using Sympy in Python

How to do symbolic differentiation in Python with SymPy

100 Most Popular Courses For September

Harvard and MIT’s $800 Million Mistake: The Triple Failure of 2U, edX, and Axim Collaborative

The future of Coursera’s only credible alternative for universities rests in the hands of 2U’s creditors.

• [2024] 1800+ Coursera Courses Still Completely Free
• 7 Best Autodesk Maya Courses for 2024: Exploring 3D Animation
• [2024] The 100 Top FREE EdX Courses of All Time
• 15 Best Lisp Courses for 2024
• 8 Best Adobe XD Courses for 2024

600 Free Google Certifications

Most common

• project management
• cyber security

Popular subjects

Communication Skills

Data Analysis

Digital Marketing

Popular courses

Greening the Economy: Sustainable Cities

Product Management Essentials

El rol de la digitalización en la transición energética

Organize and share your learning with Class Central Lists.

View our Lists Showcase

Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

Data Visualization with Python

IBM via Coursera Help

• Learn How to Sign up to Coursera courses for free
• 1800+ Coursera Courses That Are Still Completely Free
• Introduction to Data Visualization Tools
• Data visualization is a way of presenting complex data in a form that is graphical and easy to understand. When analyzing large volumes of data and making data-driven decisions, data visualization is crucial. In this module, you will learn about data visualization and some key best practices to follow when creating plots and visuals. You will discover the history and the architecture of Matplotlib. Furthermore, you will learn about basic plotting with Matplotlib and explore the dataset on Canadian immigration, which you will use during the course. Lastly, you will analyze data in a data frame and generate line plots using Matplotlib.
• Basic and Specialized Visualization Tools
• Visualization tools play a crucial role in data analysis and communication. These are essential for extracting insights and presenting information in a concise manner to both technical and non-technical audiences. In this module, you will create a diverse range of plots using Matplotlib, the data visualization library. Throughout this module, you will learn about area plots, histograms, bar charts, pie charts, box plots, and scatter plots. You will also explore the process of creating these visualization tools using Matplotlib.
• Advanced Visualizations and Geospatial Data
• Advanced visualization tools are sophisticated platforms that provide a wide range of advanced features and capabilities. These tools provide an extensive set of options that help create visually appealing and interactive visualizations. In this module, you will learn about waffle charts and word cloud including their application. You will explore Seaborn, a new visualization library in Python, and learn how to create regression plots using it. In addition, you will learn about folium, a data visualization library that visualizes geospatial data. Furthermore, you will explore the process of creating maps using Folium and superimposing them with markers to make them interesting. Finally, you will learn how to create a Choropleth map using Folium.
• Creating Dashboards with Plotly and Dash
• Dashboards and interactive data applications are crucial tools for data visualization and analysis because they provide a consolidated view of key data and metrics in a visually appealing and understandable format. In this module, you will explore the benefits of dashboards and identify the different web-based dashboarding tools in Python. You will learn about Plotly and discover how to use Plotly graph objects and Plotly express to create charts. You will gain insight into Dash, an open-source user interface Python library, and its two components. Finally, you will gain a clear understanding of the callback function and determine how to connect core and HTML components using callback.
• Final Project and Exam
• The primary focus of this module is to practice the skills gained earlier in the course and then demonstrate those skills in your final assignment. For the final assignment you will analyze historical automobile sales data covering periods of recession and non-recession. You will bring your analysis to life using visualization techniques and then display the plots and graphs on dashboards. Finally, you will submit your assignment for peer review and you will review an assignment from one of your peers. To wrap up the course you will take a final exam in the form of a timed quiz.

Saishruthi Swaminathan

• united states

Related Courses

Data visualization with r, python for data visualization: matplotlib & seaborn, crash course on interactive data visualization with plotly, python for data visualization:matplotlib & seaborn(enhanced), learn data visualization with python, related articles, from data to insights: 10 best data analysis courses for 2024, 10 best python courses for 2024: charming the snake, 1800+ coursera courses that are still completely free, 250 top free coursera courses of all time, massive list of mooc-based microcredentials.

4.5 rating at Coursera based on 11723 ratings

Select rating

Start your review of Data Visualization with Python

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

• The Softer Side of Data Visualization Training Course

Ken and I are known for our technical abilities within Tableau. I mean, out of 400 blog posts on our website, I'd say that 390 of them are directly related to Tableau. But we've been building charts and dashboards for a long, long time and have worked for, worked with, or simply collaborated alongside some of the world's foremost data visualization experts in the world. So we know a thing or two about data visualization as a practice, not just about technical aspects of Tableau. 

We are really excited to announce our first course that focuses on data visualization as a practice -  The Softer Side of Data Visualization . This course will not only dig deep into design best practices, but will also equip attendees with other essential data visualization soft skills to  create impactful data visualizations through empathy, trust-building, and data storytelling. Below are the course details.

Course Overview:  

Explore data visualization beyond the technical aspects with The Softer Side of Data Visualization  training course. This course will equip you with essential skills to create impactful data visualizations through empathy, trust-building, storytelling, and design best practices.

• Build great partnerships by asking the right questions and empathizing with stakeholders.
• Convey value over time to develop trust in both your professional relationships and the data you provide.
• Become a proactive partner, anticipating needs and offering innovative solutions.
• Employ design best practices including (but not limited to 😁) simplicity, effective use of color, data & charts that don't deceive, and selecting the right chart for the job.
• Enhance your data storytelling by focusing on the Four Ps: People, Place, Plot, and Purpose.

Who Should Attend :  

This course is designed for data professionals, business analysts, and anyone interested in improving their data visualization and storytelling skills to create more meaningful and impactful insights.

Course Format : 

This engaging 5-class-hour course is delivered in two separate sessions, approximately three weeks apart. The first session includes interactive lectures, practical exercises, and a hands-on dashboard assignment where participants apply their learning to a real-world scenario. The second session is equally valuable, as it focuses on reviewing and critiquing a selection of anonymous participant dashboards as a group.

And although this course is tool-agnostic, you all know that our favorite tool is Tableau and you will see lots of Tableau examples used throughout. 

We've already taught this course in several private settings, but we are so excited to finally make this course available publicly. If you’re interested, then please check out our  registration page .  

And, if your company is looking for private training, let us know—we can build a fully-customized training to best meet the needs of your team.

Thanks for all your support over the years! We hope to see some of you in class soon!

Kevin Flerlage ,  September 9 ,  2 024

Twitter  |  LinkedIn  |  Tableau Public

You Might Also Like

• Next You are viewing Most Recent Post

No comments:

Data Visualization & Tableau Training

We offer both generalized data visualization training and Tableau training (beginner and advanced). These are available both publicly and privately. To learn more, click the image below.

Quick Links

• Essential Reading
• Tableau & Data Visualization Training
• Moxy Analytics

Featured post

Ken and I are known for our technical abilities within Tableau. I mean, out of 400 blog posts on our website, I'd say that 390 of them a...

Popular Posts

Blog Archive

• A Walk Through My Process of Building 4 Makeover M...
• ►  August (3)
• ►  July (2)
• ►  June (2)
• ►  May (4)
• ►  April (4)
• ►  March (2)
• ►  February (4)
• ►  January (3)
• ►  December (1)
• ►  November (2)
• ►  October (4)
• ►  September (4)
• ►  August (4)
• ►  July (4)
• ►  June (3)
• ►  April (5)
• ►  March (4)
• ►  February (3)
• ►  January (4)
• ►  December (2)
• ►  November (4)
• ►  October (3)
• ►  September (3)
• ►  July (3)
• ►  April (2)
• ►  March (3)
• ►  February (5)
• ►  June (5)
• ►  May (6)
• ►  March (5)
• ►  December (3)
• ►  October (5)
• ►  September (6)
• ►  August (6)
• ►  June (4)
• ►  May (5)
• ►  April (6)
• ►  March (6)
• ►  February (6)
• ►  December (4)
• ►  November (3)
• ►  July (5)
• ►  March (8)
• ►  February (9)
• ►  January (8)
• ►  December (5)
• ►  November (5)
• ►  October (6)
• ►  August (2)
• ►  May (2)
• ►  April (3)
• ►  January (7)
• ►  October (2)
• ►  September (5)
• ►  May (3)
• ►  February (7)
• ►  November (6)
• ►  August (5)
• ►  April (1)
• ►  October (1)

Search This Blog

Report Abuse

• Big News (2)
• Data Governance & Literacy (1)
• DataViz (244)
• Design (11)
• Election (21)
• Gradient (3)
• Hire Us! (4)
• How-To (205)
• IronViz (15)
• Politics (25)
• Pop Culture (11)
• Python (10)
• Religion (6)
• Science (8)
• Series: Beyond Show Me (3)
• Series: Data Modeling (4)
• Series: Data Sources (2)
• Series: Guide to Embedding Tableau (3)
• Series: No/Yes Polygons (2)
• Series: QGIS for Tableau Users (6)
• Series: Regex (3)
• Series: Sankey Templates (4)
• Series: Set Control (3)
• Series: SQL for Tableau Users (7)
• Series: Tips (19)
• Set Control (3)
• Social Issues (22)
• Sports (21)
• Tableau (366)
• Tableau Conference (8)
• Tableau Prep (14)
• Tableau Server & Online (7)
• Templates (24)
• Timelines (2)
• Training (3)

We've joined Moxy Analytics ! To learn more or hire us, visit moxyanalytics.com .

Instantly share code, notes, and snippets.

eltomali / Data Analysis with Python Peer Graded Assignment.ipynb

• Download ZIP
• Star ( 17 ) 17 You must be signed in to star a gist
• Fork ( 21 ) 21 You must be signed in to fork a gist
• Embed Embed this gist in your website.
• Share Copy sharable link for this gist.
• Clone via HTTPS Clone using the web URL.
• Learn more about clone URLs
• Save eltomali/df20f199e764f1b289f51f6403539e29 to your computer and use it in GitHub Desktop.

pradneshbhike commented Apr 12, 2020

Sorry, something went wrong.

atere4ever commented Sep 12, 2020

sahilsaqi commented Jul 24, 2021

bro i got error in Q7! TypeError: list indices must be integers or slices, not str how to resolve this

Livhuwani-coder commented Oct 4, 2021

Thank you very much, help me a lot.

Please help me with Module 6: Databases and SQL assignment.

Thank you in advance.