Update lectures for quantecon 0.13.1

[mmcky]

quantecon 0.13.1, released today, fixes the Markov-jump Riccati recursion in LQMarkov (QuantEcon/QuantEcon.py#982 and QuantEcon/QuantEcon.py#983). This removes the notes in markov_jump_lq and policy_regimes saying that LQMarkov uses the averaged recursion. It also adds a sentence to arma saying that ma_poly and ar_poly must be padded to the same length before they are passed to dimpulse, the bug QuantEcon/QuantEcon.py#976 fixed.

The lectures install quantecon with pip install --upgrade, so the next build uses 0.13.1. Expect small changes in the markov_jump_lq figures, where decision rules move by up to 2 percent, and in the tax_smoothing_1–3 figures. The AR(2) and AR(4) impulse responses and sample paths in arma also change: before QuantEcon/QuantEcon.py#976 they were drawn too late.

🤖 Generated with Claude Code

[netlify]

✅ Deploy Preview for lustrous-melomakarona-3ee73e ready!

Name	Link
🔨 Latest commit	7a949de
🔍 Latest deploy log	https://app.netlify.com/projects/lustrous-melomakarona-3ee73e/deploys/6ad34b812d77a8ba4929272e
😎 Deploy Preview	https://deploy-preview-393--lustrous-melomakarona-3ee73e.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[Copilot]

🟢 Approval recommended

The changes are documentation-only, technically accurate, and leave no stale references to the removed LQMarkov caveats.

No open issues 🎉

What changed in this PR

This PR updates three lecture notebooks to reflect the quantecon 0.13.1 release. Since the lectures install quantecon with pip install --upgrade, the next build picks up the fixed LQMarkov Markov-jump Riccati recursion (QuantEcon/QuantEcon.py#982, #983) and the dimpulse padding fix (#976). The changes remove now-obsolete caveats and add a clarifying note, keeping the prose consistent with the upgraded library behavior.

Changes:

• Remove the note in policy_regimes.md stating that quantecon.LQMarkov averages over next-period Markov states outside the matrix inverse.
• Remove the notes in markov_jump_lq.md that LQMarkov implements the averaged (incorrect) recursion and that its decision rules differ by up to 2 percent.
• Add a sentence in arma.md explaining that ma_poly and ar_poly must be padded to the same length before being passed to dimpulse.

File	Description
lectures/​policy_regimes.md	Removes the obsolete caveat about LQMarkov's averaged recursion now that 0.13.1 fixes it.
lectures/​markov_jump_lq.md	Removes two notes about the averaged recursion and the 2% decision-rule discrepancy, leaving the still-valid mathematical statement intact.
lectures/​arma.md	Adds an explanation that the shorter polynomial is padded with trailing zeros to avoid dimpulse delaying the response.

I verified there are no dangling references to the removed notes in these files (the remaining math statement in markov_jump_lq.md:241 is implementation-independent and still valid), and confirmed the new arma.md explanation is technically accurate: scipy reads the numerator/denominator in descending powers of $z$, so padding the shorter numerator with trailing zeros cancels the relative delay introduced by a longer ar_poly.

_{Review effort: Balanced}

[mmcky]

@jstac @thomassargent30 I have just released quantecon==0.13.1 with the fix. Thanks @thomassargent30 for reporting this and thanks @jstac for filing the issue. These are some suggested edits to the lecture now that this is on PyPI and condo-forge.

[github-actions]

🚀 Deployed on https://6ad34e3330576102db6b83a7--lustrous-melomakarona-3ee73e.netlify.app