About
Credits.
Plotwise is a small interactive course on data understanding: reading a dataset one variable, two variables and many variables at a time, by playing with it.
The creator
Luca Bonaldo
Software engineer · MSc Computer Engineering & AI
Designed and built Plotwise. I build scalable systems, IoT platforms and data pipelines, and I like making technical ideas tangible: Flap Lab, PrologEZ and FumbleChess are other tools of mine that run entirely in the browser.
Datasets
The data
- Titanic: the passenger list popularized by the Kaggle competition “Titanic: Machine Learning from Disaster” (891 passengers, training set).
- Iris: R. A. Fisher's 1936 measurements of 150 iris flowers, via the UCI Machine Learning Repository.
-
Bank data (
bank-data.csv): 600 customers and whether they bought a Personal Equity Plan, the classic Weka tutorial dataset. - Eating in the UK: average grams of 17 foods per person per week in the four UK countries, from the UK government's Family Food survey, made famous by Victor Powell and Lewis Lehe's explainer on setosa.io.
- Synthetic data (skewness and kurtosis samples, ice cream and drownings, 5D blobs, citizens' answers): generated in your browser from a fixed seed.
Built with
Open source
- Vitebuild tool
- three.js3D scenes and the title screen
- D3every 2D chart
- KaTeXthe formulas
- Paper & inkthe lucabonaldo.dev design system
- Fraunces & Intertypefaces
Statistics, PCA (Jacobi eigen-decomposition) and t-SNE are implemented from scratch in plain JavaScript, so you can read exactly what each lab computes. Nothing leaves your browser: no backend, no tracking, no account.