Maths Whimsy with Python
summary
Prototypes use numpy and matplotlib to turn whimsy into testable claims: a gridded gravity model for a square planet shows the gravity vector pointing into corners rather than cliffs, and switching from loops to meshgrid yields a 30× speedup. A bike-phone pipeline extracts wheelie time series, a model predicts wheelie duration and an autoencoder generates synthetic traces. A confusion-matrix-as-Sankey visualization helps compare classifiers. Notebooks and modules live in a public repo and most experiments actually work.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:25Acknowledgement and overview
- 2:30Defining Whimsy
- 3:55Career path to Python
- 5:41Square planet gravity
- 8:08From loops to NumPy
- 10:173D simulations and shapes
- 12:25Square planet outcomes
- 13:27Wheelie data analysis
- 14:31Wheelie coach with ML
- 16:14ML performance with Sankey
- 19:24Project learnings
- 20:50Small-step learning process
- 22:36Maths Whimsy repo
- 23:37Automating puzzles
- 25:02Quantifying projects
- 27:52Python libraries benefits
- 29:18Key lessons
- 31:09Q&A
related talks
28mDerivations, Not Just Simulations: Teaching Applied Mathematics with Scientific Python
01h 00mIntroduction to Python
01h 38mBasic Python Packages for Science
49mUntitled13.ipynb
23mPython(x,y): diving into scientific Python
34mPython ate my Homework!
01h 49mBuilding Web-based Analysis & Simulation Platforms
38mCat modelling with python