Skip to main content

Practice Projects

Skill Level: Beginner to Advanced

Why Build Projects?​

Projects allow you to apply concepts learned throughout this guide.

Instead of only following step-by-step tutorials, you’ll solve real-world problems using data and turn your skills into visible work you can share.

Beginner Projects​

These projects use beginner-friendly datasets that are relatively small and well described. You can either work directly with the Kaggle files or create your own cleaned copies.

Student Performance Analysis​

Use a student performance dataset to explore questions such as:

  • How do scores differ by gender or lunch type?
  • Does test preparation seem related to performance?
  • Which subject shows the highest average scores?

Dataset on Kaggle:
Students Performance in Exams

Try to:

  • load the data with Pandas,
  • create simple summary tables,
  • make at least three charts (bar, boxplot, and scatter),
  • write 3–4 sentences describing your findings.

Coffee and Code Analysis​

Use a coffee sales or transaction dataset and imagine it as “coffee and code” sessions.

Possible questions:

  • What products are most popular?
  • Are there clear peaks in sales by time or day?
  • How might coffee availability relate to study sessions?

Datasets on Kaggle (choose one):
Coffee Sales Dataset

Try to:

  • explore basic sales patterns,
  • visualize quantities or revenue over time,
  • write a short reflection on what you notice.

Weather Data Exploration​

Pick a city and use daily climate or weather data.

Questions to explore:

  • How do temperatures change across months?
  • Which month is the warmest or coldest?
  • Are there patterns in humidity or wind?

Datasets on Kaggle (examples):
Daily Weather Data for Major Cities

Try to:

  • focus on one location and time period,
  • create line charts and histograms,
  • describe at least one trend and one unusual observation.

Movie Ratings Analysis​

Use a movie-related dataset to look at ratings patterns, genres, or review summaries.

Questions you might ask:

  • How are ratings or review scores distributed?
  • Do certain genres tend to have higher scores?
  • Are newer movies reviewed differently than older ones?

Datasets on Kaggle:
Movie Reviews Dataset (Letterboxd/Metacritic)

Aim to:

  • choose a simple subset or summary of the data,
  • use histograms and basic aggregations,
  • explain what the charts suggest about ratings or sentiment patterns.

Intermediate Projects​

Intermediate projects use richer datasets—often from public sources—which may require a bit more cleaning or selection.

Sales Dashboard​

Build a small dashboard that shows:

  • monthly sales,
  • top products,
  • basic trends over time.

Possible dataset sources:

  • Kaggle retail/sales datasets (for example, search “sales” or “retail” on Kaggle)

Include:

  • a line chart for sales over time,
  • a bar chart for product comparison,
  • a short written summary of key insights.

Spotify Listening Analysis​

Use a Spotify dataset to analyze tracks, artists, or genres.

Possible dataset source:

Goals:

  • understand most played artists or genres,
  • see patterns over years,
  • explore audio features such as danceability or energy.

Netflix Data Exploration​

Analyze a Netflix titles dataset.

Dataset example:

Questions:

  • How many titles by genre?
  • How has content changed over years?
  • Are there patterns in release countries?

World Happiness Report​

Use the World Happiness Report data to explore:

  • how happiness scores vary by region,
  • which factors (GDP, social support) seem related to happiness,
  • changes over time for selected countries.

Dataset:

Include:

  • a few comparative charts,
  • a brief interpretation of what might influence scores.

Advanced Projects​

Advanced projects can involve complex, multi-table, or time‑series data and may require more careful cleaning.

V-Dem Political Analysis​

Explore democracy indicators from the V-Dem Institute.

Primary data source (non-Kaggle, but standard for this topic):

Ideas:

  • compare democracy scores across regions,
  • study changes over time for selected countries,
  • relate democracy scores to other indicators.

Election Data Dashboard​

Use election results data to:

  • visualize vote shares by region,
  • compare parties across time,
  • highlight turnout patterns.

Dataset example on Kaggle:

Or use national/state election open data portals.

Healthcare Data Analysis​

Analyze healthcare-related datasets to study:

  • appointments,
  • outcomes,
  • regional differences.

Dataset options on Kaggle (browse and choose one):

Focus on:

  • a clear question,
  • a mix of descriptive stats and visualizations,
  • thoughtful communication of results.

Climate Change Visualization​

Use climate-related datasets such as temperature anomalies or CO₂ levels.

Possible Kaggle and external sources:

  • Climate/temperature/CO₂ datasets on Kaggle (search “climate change”, “temperature”, “CO2”)
  • Climate data from Our World in Data or NASA (outside Kaggle)

Try to:

  • build at least one long‑term trend chart,
  • annotate important events or changes.

Public Dataset Sources (General)​

You can explore many more datasets at:

  • Kaggle
  • UCI Machine Learning Repository
  • Data.gov
  • World Bank Open Data
  • Our World in Data
  • V-Dem Institute

Pick a dataset that interests you, then apply the project workflow below.

Project Workflow​

A typical data analysis project includes:

  1. Define the problem
  2. Collect the data
  3. Clean the data
  4. Explore the data
  5. Analyze the data
  6. Visualize findings
  7. Communicate results

For the beginner projects on this page, you can treat most Kaggle datasets above as your starting point, focusing mainly on exploration, analysis, visualization, and communication.

Portfolio Tip​

As you complete projects, publish them on GitHub.

A portfolio containing well‑documented projects (clear README, comments, and screenshots) shows your technical skills and analytical thinking to instructors, potential employers, and graduate programs.

Even a few small beginner projects can make a big difference if they are explained clearly and linked to datasets that others can reuse.