Quantitative Economics with Python Development

Expert guidance for working with the QuantEcon lecture-python.myst repository - intermediate quantitative economics lecture materials built with Jupyter Book.

Overview

This skill provides comprehensive instructions for developing, building, and maintaining the Intermediate Quantitative Economics with Python lecture series. The repository contains 80+ MyST markdown lecture files that are executed and compiled into HTML, PDF, and notebook formats using Jupyter Book.

Step-by-Step Instructions

1. Environment Setup

Always execute these commands in sequence before any development work:

```bash

Create conda environment (3-5 minutes - DO NOT CANCEL)

conda env create -f environment.yml

Activate environment (required for all subsequent commands)

source /usr/share/miniconda/etc/profile.d/conda.sh

conda activate quantecon

Install PyTorch with CUDA (3-5 minutes - DO NOT CANCEL)

pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128

Install additional dependencies

pip install pyro-ppl

pip install --upgrade "jax[cuda12-local]==0.6.2"

pip install numpyro

Verify installation

python scripts/test-jax-install.py

```

**Critical**: Never cancel installation commands. Each step has expected timing and interruption will require restarting from scratch.

2. Building the Project

**WARNING**: Build commands take 45-90 minutes to complete. Never cancel them.

```bash

HTML build (primary target, 45-60 minutes)

jb build lectures --path-output ./ -W --keep-going

PDF build via LaTeX (30-45 minutes)

jb build lectures --builder pdflatex --path-output ./ -n -W --keep-going

Jupyter notebook build (30-45 minutes)

jb build lectures --path-output ./ --builder=custom --custom-builder=jupyter -n -W --keep-going

```

Builds execute all notebooks sequentially, caching results in `_build/.jupyter_cache/`.

3. Making Content Changes

When editing lecture materials:

1. **Activate environment**: `conda activate quantecon`

2. **Edit MyST files**: Modify `.md` files in `/lectures/` directory

3. **Test locally**: Consider testing on a subset of changed files if possible

4. **Full validation**: Run complete HTML build to ensure all notebooks execute

5. **Verify output**: Check `_build/html/` for expected results

4. Validation

After changes, run these checks:

```bash

Verify scientific stack

python scripts/test-jax-install.py

Check environment packages

conda list | grep -E "(jax|torch|jupyter-book)"

Expected: jax 0.6.2, torch 2.9.0.dev, jupyter-book 1.0.3

View repository structure

ls -la /home/runner/work/lecture-python.myst/lecture-python.myst/

Check lecture files

ls lectures/ | head -10

```

5. Understanding Build Progress

During builds, monitor:

6. Working with Project Structure

**Key directories:**

**Technology stack:**

7. Troubleshooting

**Expected warnings (safe to ignore):**

**Performance notes:**

Important Constraints

1. **Never cancel long-running commands** - Installation and build processes must complete fully

2. **Always activate conda environment** before running any commands

3. **Expect long build times** - 45-90 minutes is normal, not an error

4. **Network warnings are expected** in sandboxed environments

5. **GPU support may be unavailable** - JAX runs in CPU mode and is still functional

Common Tasks Reference

```bash

Quick environment check

conda activate quantecon && python scripts/test-jax-install.py

Monitor build output location

ls -lh _build/html/

Check cache size

du -sh _build/.jupyter_cache/

View recent lecture modifications

ls -lt lectures/*.md | head -5

```

CI/CD Integration

The repository uses GitHub Actions with cache, CI, and publish workflows. Local builds should match CI behavior using identical commands and expecting similar timing.