Address PR #374 R5 P3: stale to_dataframe docstring + Mammen/Webb coverage

P3 #1: ``to_dataframe`` method docstring at
``chaisemartin_dhaultfoeuille_results.py:1375-1379`` listed the
pre-change ``level="by_path"`` schema (no ``cband_*`` columns) even
though the implementation now returns them. Updated the bullet to
include ``cband_lower / cband_upper``, document the negative-horizon
placebo convention, and document the NaN-on-absent-band behavior.

P3 #2: ``TestByPathSupTBands::test_path_sup_t_seed_reproducibility``
only exercised the default ``rademacher`` weight family. Parameterized
over ``["rademacher", "mammen", "webb"]`` to pin that the per-path
sup-t branch correctly threads ``self.bootstrap_weights`` through
``_generate_psu_or_group_weights`` for all three multiplier families
the feature advertises. The existing OVERALL machinery handles all
three uniformly, but the per-path surface lacked direct coverage.
Each variant must produce a finite, reproducible crit on the standard
3-path fixture.

17 tests pass on TestByPathSupTBands (was 15: +2 new parameterized
variants on the existing seed_reproducibility test).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: igerber

diff_diff/chaisemartin_dhaultfoeuille_results.py

@@ -1373,10 +1373,19 @@ def to_dataframe(self, level: str = "overall") -> pd.DataFrame:
             - ``"design2"``: Design-2 switch-in/switch-out descriptive
               summary. Available when ``design2=True``.
             - ``"by_path"``: one row per (path, horizon) when
-              ``by_path=k`` was passed to the estimator. Columns include
+              ``by_path=k`` was passed to the estimator. Columns:
               ``path``, ``frequency_rank``, ``n_groups``, ``horizon``,
               ``effect``, ``se``, ``t_stat``, ``p_value``,
-              ``conf_int_lower``, ``conf_int_upper``, ``n_obs``.
+              ``conf_int_lower``, ``conf_int_upper``, ``n_obs``,
+              ``cband_lower``, ``cband_upper``. The ``horizon`` column
+              takes negative ints for placebo rows when
+              ``placebo=True``. The ``cband_*`` columns mirror the
+              OVERALL ``level="event_study"`` schema (joint sup-t
+              simultaneous bands); they are populated for positive-
+              horizon rows of paths with a finite per-path sup-t crit
+              (``n_bootstrap > 0``) and NaN otherwise (placebo rows,
+              unbanded paths, or the requested-but-empty fallback
+              DataFrame).
 
         Returns
         -------

tests/test_chaisemartin_dhaultfoeuille.py

@@ -5661,23 +5661,54 @@ def test_path_sup_t_crit_finite_and_positive(self):
             assert entry["method"] == "multiplier_bootstrap"
             assert entry["n_valid_horizons"] >= 2
 
-    def test_path_sup_t_seed_reproducibility(self):
-        """Same seed -> bit-identical ``crit_value`` for every path."""
+    @pytest.mark.parametrize("bootstrap_weights", ["rademacher", "mammen", "webb"])
+    def test_path_sup_t_seed_reproducibility(self, bootstrap_weights):
+        """Same seed -> bit-identical ``crit_value`` for every path,
+        across all three multiplier-weight families. Pins that the
+        per-path sup-t branch correctly threads ``bootstrap_weights``
+        through ``_generate_psu_or_group_weights`` and that
+        Rademacher / Mammen / Webb each produce a finite, reproducible
+        crit (the helper handles all three uniformly under the
+        existing OVERALL sup-t machinery; this is a per-path direct
+        regression on that contract)."""
         data = _by_path_three_path_data()
         _est_a, res_a = self._fit_with_bootstrap(
-            data, by_path=3, L_max=3, n_bootstrap=200, seed=42
+            data,
+            by_path=3,
+            L_max=3,
+            n_bootstrap=200,
+            seed=42,
+            bootstrap_weights=bootstrap_weights,
         )
         _est_b, res_b = self._fit_with_bootstrap(
-            data, by_path=3, L_max=3, n_bootstrap=200, seed=42
+            data,
+            by_path=3,
+            L_max=3,
+            n_bootstrap=200,
+            seed=42,
+            bootstrap_weights=bootstrap_weights,
         )
         assert res_a.path_sup_t_bands is not None
         assert res_b.path_sup_t_bands is not None
         assert set(res_a.path_sup_t_bands.keys()) == set(res_b.path_sup_t_bands.keys())
+        # At least one path should produce a finite crit on this fixture
+        # (3 paths each with 3 valid horizons under all three weight
+        # families); pinning that the new dispatch path actually fires
+        # for `mammen` / `webb`, not just `rademacher`.
+        assert len(res_a.path_sup_t_bands) >= 1, (
+            f"bootstrap_weights={bootstrap_weights}: expected at least "
+            f"one path with a finite crit; got empty dict"
+        )
         for path in res_a.path_sup_t_bands:
             crit_a = res_a.path_sup_t_bands[path]["crit_value"]
             crit_b = res_b.path_sup_t_bands[path]["crit_value"]
+            assert np.isfinite(crit_a), (
+                f"bootstrap_weights={bootstrap_weights} path={path}: "
+                f"crit_value not finite ({crit_a})"
+            )
             assert crit_a == crit_b, (
-                f"path={path}: seed-pinned crits diverge: {crit_a} vs {crit_b}"
+                f"bootstrap_weights={bootstrap_weights} path={path}: "
+                f"seed-pinned crits diverge: {crit_a} vs {crit_b}"
             )
 
     def test_path_sup_t_skipped_when_path_has_only_one_valid_horizon(self):