calvo#

  • Series: lecture-dp

  • File: lectures/calvo.md

  • Audit date: 2026-08-26

  • Corpus snapshot: c30490a2f4

  • Categories audited: writing, math, code, figures, references, links, admonitions (JAX out of scope)

  • Overall score: 7.2 / 10

  • Priority: HIGH

Score breakdown#

Category

Score

One-line note

Writing

3/10

qe-writing-005 ×8; qe-writing-002 ×6; qe-writing-004 ×4, +4 more.

Math

5.5/10

qe-math-002 ×10.

Code

8.5/10

qe-code-001 ×4.

JAX

out of scope

JAX rules target lecture-jax.

Figures

7/10

qe-fig-003 ×2; qe-fig-001 ×5; qe-fig-008 ×1.

References

8.5/10

qe-ref-001 ×3.

Links

8/10

qe-link-002 ×4.

Admonitions

10/10

no mechanical violations detected.

Issues#

Critical#

None found.

High severity#

  • [qe-fig-001] — Do not set figure size unless necessary. Count: 5. Lines: 1051, 1352, 1390, 1414, 1447. Example: figsize=.

  • [qe-math-002] — Use \top for transpose notation. Count: 10. Lines: 272, 485, 489, 501. Example: apostrophe transpose }'.

  • [qe-writing-002] (reviewer) — Keep writing clear, concise, and valuable. Count: 6. Lines: 28, 75, 318, 725, 749, 753. Example: the installation cell appears twice: lines 28-34 (‘this lecture will need the following libraries’) and lines 77-83 (‘this lecture will use the following libraries’), running the identical !pip install --upgrade quantecon. Lines 749-750 repeat 708-709 word for word (‘Time-variation of \(\vec \mu\) chosen by a Ramsey planner is the telltale sign of the Ramsey plan’s time inconsistency’). Line 75 has ‘the recursive structure structure of the Ramsey problem’; line 753 has ‘must must choose’; lines 318 and 321 both read ‘is the value of attained by the government’; line 725 has ‘defined inequation {eq}`eq:barvdef`’. Four more of the same kind: ‘plannner’ (704), ‘for computed in subproblem 1 above’ (554), ‘and of the continuation value’ (1341), and ‘respectfully’ for ‘respectively’ (1361).

  • [qe-writing-005] (reviewer) — Use bold for definitions, italic for emphasis. Count: 8. Lines: 153, 220, 251, 329, 345, 646, 715, 753. Example: bold used for stress rather than definition at 251 (costs / benefits), 329 (policy), 345 (output / input), 646-647 (‘promised inflation equals actual inflation’), 715 (the single word is) and 753 (must); line 153 uses ‘Insight:’ as a bold pseudo-label. Line 220 is the reverse error - the term being defined, ‘bliss level’, is wrapped in double backticks and so renders as inline code rather than bold, and it is the name the lecture then relies on. The file’s genuine definitional bolds (time inconsistency 38, Ramsey plan 55, dynamic programming squared 426, promised inflation rates 634) are doing the right job and are diluted by the rest.

  • [qe-writing-008] — Remove excessive whitespace between words. Count: 242. Lines: 38, 41, 43, 45, 46, 51, 57, 59, 61, 63, …. Example: 2 spaces.

Medium severity#

  • [qe-code-001] (reviewer) — Follow PEP8 unless closer to mathematical notation. Count: 4. Lines: 962, 1259, 1297, 1455. Example: line 962 assigns a lambda (E731) and writes a space after the unary minus, lambda θ_array: - np.array([1, θ_array]), with a backslash continuation that the surrounding parentheses make unnecessary; line 1259 puts a compound statement on one line, if not isinstance(clqs, list): clqs, axes = [clqs], [axes] (E701); line 1297 runs to about 90 characters (E501); line 1455 writes T-1 unspaced where the same file spaces its other binary operators. Trailing whitespace is also scattered through the class body (923, 937, 947, 967, 998, 1001, 1008, 1010).

  • [qe-fig-003] — No matplotlib embedded titles. Count: 2. Lines: 1263, 1463. Example: .set_title.

  • [qe-link-002] — Use doc links for cross-series references. Count: 4. Lines: 67, 193, 288, 765. Example: raw link to python-intro.quantecon.org.

  • [qe-ref-001] — Use correct citation style. Count: 3. Lines: 122, 153, 445. Example: {cite} in narrative flow: ‘. {cite}’.

  • [qe-writing-001] — Use one sentence per paragraph. Count: 2. Lines: 287, 416. Example: 2 sentences in one paragraph.

  • [qe-writing-003] (reviewer) — Maintain logical flow. Count: 2. Lines: 432, 644. Example: the section that names the lecture’s method points at the wrong two equations. ‘Note on dynamic programming squared’ (423-437) says the two Bellman equations are ‘{eq}`eq_old1`, which expresses how \(\theta_t\) depends on \(\mu_t\) and \(\theta_{t+1}\)’ and ‘{eq}`eq_old4`, which expresses how value \(v_t\) depends on \((\mu_t, \theta_t)\) and \(v_{t+1}\)’. But eq_old1 (112) is the Cagan money-demand function \(m_t - p_t = -\alpha(p_{t+1}-p_t)\) and eq_old4 (189) is the state transition \(x_{t+1} = Ax_t + B\mu_t\). The equations meant are eq_old2 (136) and eq_old8 (302). Both references resolve, so nothing breaks - the reader simply lands on the wrong display twice in the four lines that explain what ‘dynamic programming squared’ means. Separately, line 644 says ‘(Here an application of the Big \(K\), little \(k\) trick is again at work.)’ - the trick has not appeared earlier in this lecture, is not explained, and is not linked.

  • [qe-writing-004] — Avoid unnecessary capitalization in narrative text. Count: 4. Lines: 67, 193, 765, 817. Example: mid-sentence ‘Control’.

Low severity#

  • [qe-fig-008] — Use lw=2 for line charts. Count: 1. Lines: 1454. Example: plot() without lw=.

  • [qe-writing-007] (reviewer) — Use visual elements to enhance understanding. Count: 1. Lines: 371. Example: ‘Three timing protocols’ (369-412) distinguishes three models along three stated dimensions - what a policymaker chooses, when it chooses, and what it assumes about how its choice affects expectations - and then delivers the nine resulting cells as three separate prose passages (373-378, 380-389, 396-412). The reader has to build the grid mentally, and the payoff of the whole lecture depends on holding it. A three-by-three table would do it in one block, and the lecture already renders LaTeX tables from pandas (generate_table, 1287-1300), so the machinery is present.

Strengths#

  • ‘Multiple roles of \(\theta_t\)’ (681-696) is the conceptual payoff of the whole construction and it is stated compactly: the same symbol is the actual inflation rate in {eq}`eq_old3`, the public’s expected rate in {eq}`eq_old2`, and a promise chosen at time 0 in {eq}`eq_old9` - and lines 692-696 name what that ambiguity illuminates (whether the government leads or follows the market, forward guidance, inflation targeting).

  • Chang’s simplifying insight is given as a reason rather than a step: lines 153-157 explain that \(\theta_t\) intermediates how every future \(\mu_{t+j}\) reaches time-\(t\) real balances, so an equivalence class of continuation money-growth sequences delivers the same \(\theta_t\) - which is exactly what licenses treating \(\theta\) as the state in the sections that follow.

  • The comparative statics are predicted before they are shown: 1367-1377 derives from the closed forms that changing \(\beta\) must move \(\theta_\infty^R\) and \(\theta_0^R\) but leave \(\theta^*\), \(\theta^{CR}\) and \(\theta^{MPE}\) alone, 1401-1405 says which of those change when \(c\) moves instead, and 1421-1424 explains the \(c \to 0\) limit by naming Calvo’s wedge. The figures then confirm claims the reader has already been given.

  • Every model in the lecture is an attribute of one ChangLQ instance built by six methods called in a documented order (908-918), so the Ramsey, constrained-Ramsey and Markov-perfect numbers that appear side by side in the tables are guaranteed to come from the same parameter set rather than from three separate calibrations.

  • The five \(\theta\) landmarks get their colours and labels from a single helper, compute_θs (1029-1043), which every figure and every table calls - so \(\theta^*\), \(\theta_\infty^R\), \(\theta^{CR}\), \(\theta_0^R\) and \(\theta^{MPE}\) keep the same colour and the same label across the whole lecture, including across the parameter sweeps at 1352, 1390 and 1414.

  • The Ramsey problem is split into two explicitly numbered subproblems (462, 573), the split is justified by reference to {doc}`dyn_stack` and Ljungqvist-Sargent chapter 19 (445), and subproblem 2 is carried through to a closed form for \(\theta_0^R\) (600-610) rather than handed to a solver.

  • The {note} at 415-420 says precisely which sibling lecture handles the fourth timing protocol ({doc}`calvo_abreu`), what distinguishes it, and which literature it belongs to - so a reader who wonders about credible policy is not left guessing.