feat(v4): item 9b of the 4.0 cut - retire the 5.0-window shims at 4.0 + placebo wrappers take the calendar-time contract (M-016, M-050..M-058, M-148, M-149, M-151, M-152 removed; M-156, M-157)

TODO.md

@@ -41,7 +41,6 @@ Related tracking surfaces:
 | ContinuousDiD post-fit bootstrap-ES REPLAY: since 4.0 PR 3c a bootstrapped fit relays the event-study surface computed at fit (M-025), so the gap is a genuine post-fit replay (fresh draws, no fit-side cost, a route for pre-4.0 pickles); a seeded replay is tractable - the multiplier draws are seeded (`np.random.default_rng(self.seed)`) - but needs the FULL per-cell `_bootstrap_info` (bread/ee_treated/Psi_eval/dPsi_*/beta_pred) the pruned kit deliberately drops, so shipping it means a kit-payload change with its own memory contract | `diff_diff/continuous_did_aggregation.py`, `diff_diff/continuous_did_results.py` | 2(b) PR-3c | Mid | Low |
 | practitioner `step_name="heterogeneity"` producer-side collisions: three OTHER estimators' advice steps reuse the key with non-heterogeneity labels (`:975` ContinuousDiD dose-response, `:1022` Triple placebo-group, `:1413` LPDiD WAS arrays), so DiagnosticReport's heterogeneity completion silently drops that unrelated advice from `next_steps` via `_filter_steps` - the same latent collision fixed for StackedDiD in M-024 (renamed to `sub_experiment_balance`). Renaming these changes those estimators' report output; audit + rename with per-estimator pins. | `diff_diff/practitioner.py` | 2(b) PR-2 review R9 | Quick | Low |
 | PreTrendsPower `violation='linear'` on CS `base_period='varying'` input targets the wrong alternative: `δ_pre = M · \|t\|` assumes level coefficients against a common reference, but varying-base pre-treatment effects are consecutive-period comparisons (constant increments under a linear trend). Both CS-sourced routes now WARN (REGISTRY PreTrendsPower Note), and universal-base GAPPED grids fail closed via the `reference_event_times` common-reference guard; what remains is the varying-base resolution - either transforming the violation vector through each coefficient's actual base mapping (needs per-horizon base provenance) or requiring `base_period='universal'` for the linear benchmark - a per-estimator methodology decision with a hand-calculated linear-violation gate | `diff_diff/pretrends.py` | 2(b) PR-1 R5 | Mid | Medium |
-| `run_placebo_test`/`run_all_placebo_tests` expose ONE overloaded `time` param (calendar for timing/group tests, 0/1 post dummy forwarded into the M-137/M-138 callees - works only because a two-period 0/1 calendar column is both); redesign the wrapper signatures (dual `time=`/`post=` params or a split) so each callee gets an honestly-named argument | `diff_diff/diagnostics.py` | 2(c)-ii PR-B | Mid | Medium |
 | `absorbed_fe_rank` N>=3 general rank: the helper keeps `sum(levels-1)` for 3+ absorbed dims, exact for independent connected dims but an over-count for duplicated/nested triples (measured `a(5),b(4),c==b(4)`: true 7 vs formula 10) and for disconnected N-way graphs. Two-way is component-exact. Deriving general N-way FE rank is a hypergraph problem; do it with a reference (fixest's `fixef.rm`/reghdfe df_a) rather than a guess. See `docs/methodology/variance-conventions.md` D3. | `diff_diff/utils.py` | #variance-inventory | Mid | Low |
 | `SyntheticControl` conformal (CWZ 2021) AR / innovation-permutation path (Lemmas 5-7) for time-series proxies — the residual-permutation shortcut is only valid for time-permutation-invariant proxies (SC/Lasso/DiD); an AR proxy needs innovation permutation. | `diff_diff/conformal.py`, `diff_diff/synthetic_control_results.py` | CWZ-2021 | Heavy | Low |
 | Re-run the R-dependent benchmark refresh so `docs/benchmarks.rst`'s TWFE "SE Rel Diff 0.1%" cell reflects the 3.9 K_reference convergence (expected 0.1% -> 0.0%; the table is generated, never hand-edited — the movement is noted in the CHANGELOG entry). | `docs/benchmarks.rst`, `benchmarks/R/` | #variance-inventory | Quick | Low |

benchmarks/speed_review/bench_reversible_dcdh.py

@@ -69,7 +69,7 @@ def inspect_placebo():
         r = results["dcdh"]
         results["placebo_summary"] = {
             "placebo_effect": getattr(r, "placebo_effect", None),
-            "overall_att": getattr(r, "overall_att", None),
+            "att": getattr(r, "att", None),
             "joiners_att": getattr(r, "joiners_att", None),
             "leavers_att": getattr(r, "leavers_att", None),
         }

changelog.d/20261018-v4-cut-framing.md

@@ -24,6 +24,6 @@
   keeps due enforcement unchanged. The id snapshot and `ROW_COUNT_FLOOR` now include
   `M-148`/`M-149` (shipped by #838 without extending them) and `M-150` (132 rows).
   `M-091`'s roster records the four later-born `Diagnostic` containers; `M-050`'s
-  `code_refs` record that `diff_diff/_deprecation.py` survives 4.0 as 5.0-window
+  `code_refs` record that `diff_diff/_deprecation.py` survives 4.0 as pickle-migration
   infrastructure; `M-045`/`M-115` name the legacy-CR1 remap branches their removal
   deletes. A `maint/3.13` branch is cut at `v3.13.0` for 3.13.x patches.

changelog.d/20261018-v4-pr1-conformance.md

@@ -10,13 +10,11 @@
 - **`RDDensityTestResult.to_dict()`**: JSON-serializable scalars with native casts (NaN
   inference passes through), the `*_conventional` block `None` unless `report_all=True`, and
   the bandwidth-selection table nested under `bandwidths` only when it was estimated.
-- `diff_diff._deprecation.deprecated_field_property` takes `removed_in=` so the warning and the
-  generated docstring name the row's actual removal release.
 
 ### Deprecated
 - **`LWDiDResults.pvalue` and `LWDiDResults.ci`** (ledger rows `M-151`, `M-152`): the
-  3.10-born alias properties for `p_value` / `conf_int` now emit a `FutureWarning` and are
-  removed in 5.0. Read `.p_value` / `.conf_int`.
+  3.10-born alias properties for `p_value` / `conf_int` were first shimmed with a `FutureWarning`
+  and then removed outright at 4.0 (item 9b of the cut). Read `.p_value` / `.conf_int`.
 
 ### Breaking Changes
 - **`LWDiDResults.to_dataframe()` is reshaped onto the shared results schema** (ledger row

changelog.d/20261018-v4-pr2-shim-removals.md

@@ -55,5 +55,6 @@
 - `tests/test_v4_matrix.py` removed-status probes: a function row that resolves to a submodule
   counts as absent; a param row absent from the explicit signature counts as removed despite
   `**kwargs`. Naming-guard allowlist extended for the kept body-local aliases, `__setstate__` keys
-  and shared vocabulary. The `diff_diff._deprecation` helpers survive for the remaining
-  M-031/M-082, PR-3 and 5.0-window shims.
+  and shared vocabulary. The `diff_diff._deprecation` helpers survived this PR for the remaining
+  M-031/M-082 and PR-3 shims (item 9b later reduced the module to its pickle-migration and
+  required-arg helpers).

changelog.d/20261025-v4-pr4c-ii-remove-mpd.md

@@ -20,8 +20,9 @@
   (`M-083`). Through 3.x `time=` was the deprecated alias for the 0/1 post indicator and
   emitted a `FutureWarning` (`M-082`); since 4.0 `time` means the event-study calendar column
   only. Supplying both `post=` and `time=` keeps raising the "pass only post=" error.
-- **`EventStudyResults.period_effects` is a `FutureWarning` compatibility view** (`M-016`;
-  removed in 5.0): on a calendar surface it returns the former calendar-keyed dict of
+- **`EventStudyResults.period_effects` is removed** (`M-016`; the compatibility view
+  this item first shipped was retired at 4.0 by item 9b - no release carried it). Through
+  the pre-release window, on a calendar surface it returned the former calendar-keyed dict of
   per-period records (`period, effect, se, t_stat, p_value, conf_int, is_significant,
   significance_stars`, reference period omitted; `is_significant` is evaluated at the fit's
   `alpha`, where the removed record hard-coded 0.05; duplicate labels raise). Relative-time

changelog.d/20261026-v4-pr5b-storage-flip.md

@@ -1,8 +1,8 @@
 ### Breaking Changes
 
-- **4.0 cut, storage flip of the headline quintet (`M-050`..`M-058`).** `CallawaySantAnnaResults` (and `DMLDiDResults`), `SunAbrahamResults`, `ImputationDiDResults`, `TwoStageDiDResults`, `StackedDiDResults`, `EfficientDiDResults`, `WooldridgeDiDResults`, `ChaisemartinDHaultfoeuilleResults` and `ContinuousDiDResults` now store the canonical `att` / `se` / `t_stat` / `p_value` / `conf_int` as their native dataclass fields (renamed in slot, so the positional constructor layout is unchanged) and take those names as constructor keywords (`overall_att=` raises `TypeError`). The 3.x `overall_att` / `overall_se` / `overall_t_stat` / `overall_p_value` / `overall_conf_int` names - and on `ContinuousDiDResults` all nine legacy spellings, the `overall_att_*` fields and the former silent `overall_se`-style aliases - are read-only properties that emit a `FutureWarning` and are removed at 5.0; they are absent from `dataclasses.fields()`, `asdict()` and `to_dict()`. 3.x pickles load unchanged (each container's `__setstate__` rewrites the legacy keys on load). The bootstrap draw containers (`bootstrap_results.overall_att_se` and siblings) are unchanged, and `StaggeredTripleDiffResults` keeps its `overall_*` storage until the DDD container unification.
+- **4.0 cut, storage flip of the headline quintet (`M-050`..`M-058`).** `CallawaySantAnnaResults` (and `DMLDiDResults`), `SunAbrahamResults`, `ImputationDiDResults`, `TwoStageDiDResults`, `StackedDiDResults`, `EfficientDiDResults`, `WooldridgeDiDResults`, `ChaisemartinDHaultfoeuilleResults` and `ContinuousDiDResults` now store the canonical `att` / `se` / `t_stat` / `p_value` / `conf_int` as their native dataclass fields (renamed in slot, so the positional constructor layout is unchanged) and take those names as constructor keywords (`overall_att=` raises `TypeError`). The 3.x `overall_att` / `overall_se` / `overall_t_stat` / `overall_p_value` / `overall_conf_int` names - and on `ContinuousDiDResults` all nine legacy spellings, the `overall_att_*` fields and the former silent `overall_se`-style aliases - are removed at 4.0 (item 9b retired the `FutureWarning` property shims this item first shipped - reading one raises `AttributeError`); they are absent from `dataclasses.fields()`, `asdict()` and `to_dict()`. 3.x pickles load unchanged (each container's `__setstate__` rewrites the legacy keys on load). The bootstrap draw containers (`bootstrap_results.overall_att_se` and siblings) are unchanged, and `StaggeredTripleDiffResults` keeps its `overall_*` storage until the DDD container unification.
 - **Report schemas.** `DiagnosticReport.to_dict()` reports `headline_metric.name` and `target_parameter.headline_attribute` as `"att"` for the nine containers above (`"overall_att"` under 3.x); `DIAGNOSTIC_REPORT_SCHEMA_VERSION` 2.0 -> 3.0. `BusinessReport.to_dict()` reports `target_parameter.headline_attribute` likewise; `BUSINESS_REPORT_SCHEMA_VERSION` 3.0 -> 4.0 (its `headline` block carries no name key). `StaggeredTripleDiffResults` keeps `headline_attribute: "overall_att"` until item 6.
 
 ### Internal
 
-- Ledger rows `M-050`..`M-058` flip to `shimmed` with `tests/test_v4_storage_flip.py` as their `test_ref`; `diff_diff._deprecation` gains `LEGACY_OVERALL_QUINTET` and `migrate_renamed_state_keys` (5.0-window infrastructure); `DMLDiDResults` re-declares the five shims so the warning names the subclass (naming-guard allowlist reason reworded); `_extract_scalar_headline`'s legacy alias row is deleted; `tests/test_result_aliases.py` pins the post-flip direction (SDDD kept in the 3.x direction); the API autosummary stubs are regenerated; the bootstrap draw containers' payload names are recorded as a REGISTRY `Note` (user decision 2026-10-26); spec item 5 carries its SHIPPED note.
+- Ledger rows `M-050`..`M-058` flip to `shimmed` with `tests/test_v4_storage_flip.py` as their `test_ref`; `diff_diff._deprecation` gains `LEGACY_OVERALL_QUINTET` and `migrate_renamed_state_keys` (the pickle-migration infrastructure); `DMLDiDResults` re-declares the five shims so the warning names the subclass (naming-guard allowlist reason reworded); `_extract_scalar_headline`'s legacy alias row is deleted; `tests/test_result_aliases.py` pins the post-flip direction (SDDD kept in the 3.x direction); the API autosummary stubs are regenerated; the bootstrap draw containers' payload names are recorded as a REGISTRY `Note` (user decision 2026-10-26); spec item 5 carries its SHIPPED note.

changelog.d/20261101-v4-pr9b-shim-retirement-placebo-time.md

@@ -0,0 +1,52 @@
+### Breaking Changes
+
+- **4.0 cut, item 9b: the compatibility shims scheduled for a 5.0 removal are removed at 4.0** (user
+  decision - no 4.x debt; none of these `FutureWarning` shims ever reached a release, so the released
+  3.x names break without a warned release - `docs/migration-4.0.md` is the notice). The 3.x
+  `overall_att` / `overall_se` / `overall_t_stat` / `overall_p_value` / `overall_conf_int` names on
+  `CallawaySantAnnaResults` (and `DMLDiDResults`), `SunAbrahamResults`, `ImputationDiDResults`,
+  `TwoStageDiDResults`, `StackedDiDResults`, `EfficientDiDResults`, `WooldridgeDiDResults`,
+  `ChaisemartinDHaultfoeuilleResults` and - all nine spellings, `overall_att_*` included -
+  `ContinuousDiDResults` raise `AttributeError`; read the canonical `att` / `se` / `t_stat` /
+  `p_value` / `conf_int` (`M-050`..`M-058`). 3.x pickles still load (the `__setstate__` key migration
+  stays). `EventStudyResults.period_effects` is gone; read `to_dataframe()` or the columnar arrays
+  (`M-016`). `LWDiDResults.pvalue` / `.ci` are gone; read `.p_value` / `.conf_int` (`M-151`, `M-152`).
+- **`SyntheticControl(inner_max_iter=, inner_min_decrease=)` are removed** (`M-148`, `M-149`; the
+  `DeprecationWarning` shipped in 3.13.0): passing either is a `TypeError`, and the positional
+  parameters after them (`standardize`, `alpha`, `seed`, `v_cv_t0`) move two slots left - pass
+  them by keyword.
+- **`run_placebo_test` / `run_all_placebo_tests`: `time` is the calendar column only and
+  `post_periods` is required for every `test_type`** (`M-156`, `M-157`; spec section 8 rule 1). The
+  0/1 post dummy the permutation and leave-one-out tests need is derived as
+  `data[time].isin(post_periods)`; a missing `time` column, missing/empty `post_periods`, missing
+  values in `time`, `post_periods` values absent from the column, or a pre-existing `_placebo_post`
+  column raise `ValueError` (on the suite wrapper before any test runs). Calls that passed the 0/1
+  indicator as `time=` with `post_periods=[1]` return identical numbers; calls that omitted
+  `post_periods` must add it.
+
+### Fixed
+
+- `run_all_placebo_tests` and `run_placebo_test(test_type="permutation" | "leave_one_out")` no
+  longer fail on a multi-period calendar column: through 3.x the wrappers forwarded `time` itself as
+  the 0/1 post dummy, so `DifferenceInDifferences.fit` rejected it and the suite silently reported
+  the two tests as error dicts (tutorial 04's placebo cell printed `ERROR` lines). The dummy is now
+  derived from `post_periods`.
+
+### Internal
+
+- Ledger rows `M-016`, `M-050`..`M-058`, `M-151`, `M-152` flip to `removed` at 4.0 with
+  `deprecated_in: null` (no release warned); `M-148`/`M-149` flip to `removed` with
+  `deprecated_in: 3.13.0`; new behavior rows `M-156`/`M-157` (`tests/test_diagnostics.py`); the
+  ledger header, spec ladder, section 5/8/9/11 notes and the phase-5 roster carry dated amendments.
+  `diff_diff._deprecation` is reduced to the pickle-migration (`LEGACY_OVERALL_QUINTET`,
+  `migrate_renamed_state_keys`, `migrate_event_study_reference_sentinel`) and required-arg
+  (`NOT_SUPPLIED`, `require_arg`) helpers - `deprecated_field_property`, `resolve_renamed_kwarg`,
+  `warn_deprecated_kwarg` and `deprecated_kwarg_message` are deleted with their last consumers.
+  With the descriptors gone, assigning a legacy name to a results instance creates a shadow
+  attribute instead of raising (the non-slotted dataclasses) - the naming guard's consumer sweep
+  is the in-repo defense. Naming guard: the `DMLDiDResults.overall_att` surface entry and the
+  `_deprecation.py` `placebo_effects` consumer entry are deleted, the two placebo wrappers join the
+  rule-1 calendar `time` inventory, and the retired token families' remaining in-repo hits
+  (body-local `overall_att` names, the kept `__setstate__` key maps, methodology history) are
+  allowlisted. API autosummary stubs regenerated; TODO.md's placebo-wrapper redesign row closed;
+  tutorial 17 re-executed (its stored practitioner snippet echoed the old `time='post'` advice).