Learn by working with real data
Data Science Python Playground is a local-first teaching lab created by Elijah Ang. It turns data questions into editable Python workflows, visible evidence, and machine-learning decisions you can inspect step by step.
What it does
Explore bundled datasets, edit and run Python cells, build charts, investigate statistical questions, compare models, and protect the final test set while learning from realistic data.
How the lessons are made
Guided lessons use small synthetic tables so one change is easy to see. The open workspaces use bundled datasets with their sources and licences listed in Credits. Each exercise asks you to read the output or explain a decision, rather than treating a passing code check as proof that an analysis is sound.
How to use the results
Start with the question and the meaning of each row. Inspect missing values, units and the limits of the dataset before drawing a conclusion. In a modelling task, compare with a simple baseline on training data and reserve the test set for a final check. A result from a teaching example is a prompt to investigate further, not a claim about every population.
How your work is handled
Machine Learning code drafts and bundled dataset text save automatically on this device for each setup. Data and Statistics open fresh notebooks; export code you want to keep before leaving. Python variables and results are temporary and reset when you reopen the workspace. CSV and chart exports are created only when you choose them through the browser or native share sheet. There is no account, cloud synchronisation, analytics or project-operated storage for analysis work.
Web and app
The GitHub Pages site remains the computer experience. The iPhone and iPad app packages the same tested web source inside a native shell.
Support
Email elijahang77@gmail.com or call +65 8822 7539.