Lecture Improvements Project (Modernization, NumPy -> JAX conversion, AI Workflows)

[mmcky]

Description

This project will review all quantecon lectures to make improvements, check for typos and review mathematical notation, modernise code (style and function) such as migrating much of our numpy and numba code to make use of the more computationally efficient library jax[cpu,gpu].

This project is the core focus for QuantEcon over the coming months and consists of three areas of work:

1. Review and modernize lectures and code
2. Use AI tools to improve quantecon workflows
3. Translation project English to Chinese

More details for each of these can be found below, in addition to useful resources.

Some metrics for success:

1. Solve and address all Issues across our lecture repositories to Issue Trackers. Tracker Metrics
2. Tick off all assignments in tracking & task issues (linked below in each project)

Timeline: August 2025 to October 2025

Important

1. Please make sure you're using an environment that uses anaconda=2025.06 and python=3.13
2. Ensure you have access to GitHub resources (coordinate access with @mmcky)

Modernize and Review Lectures + Code

Reviewers
@HumphreyYang
@mmcky

This project includes:

1. review lecture for typos, and apply the QuantEcon style guide
2. errors in mathematical notation and/or corrections
3. add exercises to lectures that do not have them
4. make use of sphinx-proof and sphinx-exercise directives to modernise the text layout (see style guide)
5. check all links on a page to make sure they work and target the correct resource
6. (code) reviewing current code to ensure no SyntaxWarnings, FutureWarnings, DeprecationWarnings
7. (code) ensure lecture is PEP8 and QuantEcon Style Guide Compliant
8. (code) convert relevant code from numpy, numba to make use of jax[cpu] and/or jax[gpu] (Notes: this will require reading and thinking through the context of each lecture to adjust comments around timing comparisons, and discussion around how the code is written, undertaking timing comparisons and documenting them (so we can write a blog post about the improvements)).

Important

For changes to https://github.com/QuantEcon/lecture-python-intro and https://github.com/QuantEcon/lecture-python.myst please
coordinate changes with the Translation Team (see Translation Project below) via Project Tagging.

Software Baseline: anaconda=2025.06, python=3.13

Tracking & Task Coordination	Status	Target
#173	Open	Intermediate Lectures

A good example of a well put together Issue and PR:

1. [orth_proj] Suggestions lecture-python-advanced.myst#216
2. [org_proj] review and update lecture-python-advanced.myst#217

Here are some good examples of code updates for reference:

1. [inventory_dynamics] Code updates to inventory lecture lecture-jax#224
2. [mccall_model_with_separation] Conversion to JAX lecture-python.myst#501 (this PR includes an example of changes, and subsequent discussions)

Important

We are currently writing a style guide update to include specific style changes around the use of NamedTuples etc. to assist
with conversion to JAX.

AI Workflows (Issue closure, code reviews)

Members
@HumphreyYang
@mmcky
@longye-tian

Note

If you have a particular interest in AI then please get in touch with @mmcky to discuss joining this team.
This team will experiment with a range of AI tooling and will write up a guide on how to use AI for the broader team.

This project is to document how best we can use AI to help with productivity improvements, in addition to getting suggestions for making sensible code improvements. A good example is when @oyamad used ChatGPT to review some of the code in our intermediate python lecture series and opened the following issue. What is excellent about this example is it is not a copy-paste from ChatGPT but rather a synthesis of a core suggestion that was discovered while using the tool. This is a great example of the types of discovery AI tools can enable.

The AI project includes:

1. Using AI tools to improve QuantEcon workflows and procedures
2. Use AI tools to review PRs (GitHub Copilot, Claude Code)
3. Use AI tools to review code and lectures and distill useful suggestions into actionable issues
4. Use AI tools to provide initial suggestions on how to solve GitHub issues

It is important to remember, AI is a companion tool, not a final decision maker. There is some discussion of how to use AI here

Tracking & Task Issues	Status
QuantEcon/lecture-python.myst#463	Open

Translation Project

Coordinator
@nisha617

We have used Claude AI to translate our English lectures into Mandarin. Reviewers will take the translated lectures and review them to improve the quality of the translation. This includes improvements to language, applying the quantecon glossary to harmonise terms used across lectures, thinking through QuantEcon processes to make suggestions on how to improve management of multilingual lecture sites.

Tracking & Task Issues	Status
QuantEcon/lecture-python.zh-cn#21	Open

Repos
https://github.com/QuantEcon/lecture-python.zh-cn
https://github.com/QuantEcon/lecture-intro.zh-cn

---

Communications and PR Reviews

For PR comments and reviews please tag @mmcky and @HumphreyYang.

Please DO NOT tag @jstac or @thomassargent30.

We are implementing a review system, that we hope will make it more rewarding and time efficient for everyone. Humphrey and I can review the PRs and provide feedback to the team. Keep an eye on each section above as we may add additional reviewers for different parts of the project as we go along.

The reviewer team will process and collect PRs and escalate them to @jstac and @thomassargent30 on a fortnightly basis.

[mmcky]

Agenda: 29th September 2025

Time: 2:00pm to 3:00pm

1. Weekly status updates
2. Updates to manual.quantecon.org (greek unicode symbols, jax conversions)
3. Using copilot as an AI companion on GitHub
4. Questions

Member	Hours
@mmcky	Monday, 2:00pm to 3:00pm
@mmcky	Thursday, 2pm to 3pm