Codex re-review round 5: clear cluster_name on TWFE Conley path

P1 (Maintainability): TWFE drops its auto-unit-cluster on the Conley path
(`_conley_cluster_override = None` in the LinearRegression call) but still
recorded `_twfe_cluster_label = unit` in the result metadata. Downstream
consumers reading `res.cluster_name` or `res.to_dict()["cluster_name"]`
were told the SEs were CR1-clustered when they were actually Conley
spatial HAC with no clustering.

Fix: when `_fit_vcov_type == "conley"`, set `_twfe_cluster_label = None`
so result-level provenance mirrors the actual cluster IDs passed to
LinearRegression.

Regression: `test_twfe_conley_cluster_name_is_none` asserts both
`res.cluster_name is None` and that `to_dict()` doesn't advertise the
`cluster_name` key on a TWFE Conley fit.

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

Author: igerber

diff_diff/twfe.py

@@ -561,9 +561,16 @@ def _refit_twfe(w_r):
         # `self.cluster is None` AND the vcov family is cluster-compatible.
         # One-way families (`classical`, `hc2`) disable the auto-cluster (see
         # the `cluster_var` block above); report None so summary() labels the
-        # one-way family instead of "CR1 cluster-robust at unit".
-        if self.cluster is not None:
-            _twfe_cluster_label: Optional[str] = self.cluster
+        # one-way family instead of "CR1 cluster-robust at unit". The Conley
+        # path drops the auto-cluster too (`_conley_cluster_override = None`
+        # above); mirror that here so the variance-provenance metadata
+        # (`res.cluster_name`, serialized `to_dict()["cluster_name"]`)
+        # accurately reflects the cluster IDs actually passed to
+        # `LinearRegression` — i.e. None on the Conley path.
+        if _fit_vcov_type == "conley":
+            _twfe_cluster_label: Optional[str] = None
+        elif self.cluster is not None:
+            _twfe_cluster_label = self.cluster
         elif cluster_var is None:
             # One-way family path: auto-cluster was intentionally dropped.
             _twfe_cluster_label = None

tests/test_conley_vcov.py

@@ -1484,6 +1484,26 @@ def test_twfe_conley_to_dict_carries_lag_cutoff(self, panel):
         assert d["vcov_type"] == "conley"
         assert d["conley_lag_cutoff"] == 1
 
+    def test_twfe_conley_cluster_name_is_none(self, panel):
+        """Closes Codex re-review round 5 P1 (Maintainability): TWFE drops
+        its auto-unit-cluster on the Conley path (`_conley_cluster_override =
+        None`), so the variance-provenance metadata must reflect that. The
+        result's `cluster_name` is None and `to_dict()` does not advertise
+        `cluster_name` — otherwise downstream consumers would be told the
+        SEs were CR1-clustered when they're actually Conley spatial HAC.
+        """
+        from diff_diff import TwoWayFixedEffects
+
+        res = TwoWayFixedEffects(
+            vcov_type="conley",
+            conley_coords=("lat", "lon"),
+            conley_cutoff_km=2000.0,
+            conley_lag_cutoff=1,
+        ).fit(panel, outcome="y", treatment="treated", time="time", unit="unit")
+        assert res.cluster_name is None
+        d = res.to_dict()
+        assert "cluster_name" not in d
+
     def test_twfe_conley_non_numeric_time_fails(self, panel):
         """TWFE's `_treatment_post = treated * time` design step requires
         numeric `time`. Non-numeric labels (datetime64, pd.Period, strings)