numfocus / YouTubeVideoTimestamps

Adding timestamps to NumFOCUS and PyData YouTube videos!
https://www.youtube.com/c/PyDataTV
MIT License
75 stars 19 forks source link

James Powell - Are generator-coroutines really the answer? | PyData London 2024 #214

Open xames3 opened 3 weeks ago

xames3 commented 3 weeks ago

Timestamps for titled video.

Contents:

00:00 Introduction 00:31 Disclaimer about the talk 00:46 What is a function? 01:13 What is a generator? 01:50 Function returning a generator 02:12 Generator Coroutine Protocol 02:58 Iterator v/s Iterable? 03:28 Traditional python coroutine 03:57 Modern python coroutine (with async def) 04:33 Yield in an async function? 04:57 Let's talk about generators 06:00 Decomposing a computation using traditional functions 07:40 Decomposing a computation using generators 08:46 Eager evaluation v/s Lazy evaluation (in-depth) 12:22 Function to Generator to Coroutine 13:17 Nature of decomposition (Homogeneity v/s Heterogeneity) 16:08 Modeling heterogeneity using Pandas with a trading example 19:18 Iteration helpers with progressive improvements 22:33 Iteration helpers with itertools 24:28 Value disintermediation and their minimizing strategies 27:14 Expanding upon Iteration helpers with more itertools 27:43 Generators in modeling State Machines 32:35 Gotchas while using a generator 34:39 Eliminating modalities using "pumping" and their caveats 37:22 Pumping Coroutine-Coroutine model? 38:25 Revisiting Value disintermediation with an example and final notes