Building Web-based Analysis & Simulation Platforms
summary
Live code shows a working pipeline: a Numpy Monte Carlo integrator (used to approximate π, ~10 million samples ~10s) is wrapped as a Python script, exposed over a Flask JSON API, dispatched to a background worker via Celery + Redis job queue, and surfaced in a React single-page app that polls task status; optional Bokeh visualization renders points. Source code and exact build commands are provided so the stack can be reproduced end to end.
AI-written summary · about
chapters (24)
- 0:00Welcome!
- 2:29Getting the source code from Dropbox
- 12:00Algorithm for calculating Pi Digits
- 17:40Generalising the problem and serving from HTTP
- 21:30Packaging to the backend code
- 30:38Testing first version
- 32:40Exposing the algorithm to a flask server
- 38:48Testing the server with httpie
- 42:14Concurrent requests
- 44:02How celery and Redis solve that problem
- 46:00Installing the last required packages
- 49:15Running Redis
- 49:35Writing the worker code
- 53:35Running the celery worker
- 57:09Exposing tasks management in the backend code
- 1:00:30Pause, letting people catchup
- 1:08:40About doing talks at PyData
- 1:10:40Opinions on JavaScript and React build chain
- 1:13:45Explaining what is in the Dropbox frontend folder and tooling choice
- 1:21:30Writing React code
- 1:26:42Running the frontend, implementing the main component
- 1:39:33Finishing the HTML form
- 1:40:25Wiring the frontend to the backend
- 1:47:43Results from the frontend application
related talks
01h 18mKeynote
01h 00mFrom notebook to production using
32mEnabling real-time insights through stream processing in Python
01h 39mPyScript - Writing a Python application in the browser
39mHow to build stunning Data Science Web applications in Python with Taipy
59mBig data beautiful visualization on the browser with Bokeh
42mPyScript & Data Science- PyData stack on the Browser
44mUsing Python with Hadoop to create production ready big data applications