diff-diff 1.2.0__py3-none-any.whl → 1.2.1__py3-none-any.whl

diff_diff/__init__.py

@@ -98,7 +98,7 @@ from diff_diff.visualization import (
     plot_sensitivity,
 )
 
-__version__ = "1.2.0"
+__version__ = "1.2.1"
 __all__ = [
     # Estimators
     "DifferenceInDifferences",

diff_diff/staggered.py

@@ -1603,30 +1603,25 @@ class CallawaySantAnna:
                         weights * bootstrap_atts_gt[b, gt_indices]
                     )
 
-        # Compute bootstrap statistics
-        # ATT(g,t) statistics
+        # Compute bootstrap statistics for ATT(g,t)
         gt_ses = {}
         gt_cis = {}
         gt_p_values = {}
 
         for j, gt in enumerate(gt_pairs):
-            original_effect = original_atts[j]
-            boot_dist = bootstrap_atts_gt[:, j]
-
-            se = float(np.std(boot_dist, ddof=1))
-            ci = self._compute_percentile_ci(boot_dist, self.alpha)
-            p_value = self._compute_bootstrap_pvalue(original_effect, boot_dist)
-
+            se, ci, p_value = self._compute_effect_bootstrap_stats(
+                original_atts[j], bootstrap_atts_gt[:, j]
+            )
             gt_ses[gt] = se
             gt_cis[gt] = ci
             gt_p_values[gt] = p_value
 
-        # Overall ATT statistics
-        overall_se = float(np.std(bootstrap_overall, ddof=1))
-        overall_ci = self._compute_percentile_ci(bootstrap_overall, self.alpha)
-        overall_p_value = self._compute_bootstrap_pvalue(original_overall, bootstrap_overall)
+        # Compute bootstrap statistics for overall ATT
+        overall_se, overall_ci, overall_p_value = self._compute_effect_bootstrap_stats(
+            original_overall, bootstrap_overall
+        )
 
-        # Event study statistics
+        # Compute bootstrap statistics for event study effects
         event_study_ses = None
         event_study_cis = None
         event_study_p_values = None
@@ -1637,16 +1632,14 @@ class CallawaySantAnna:
             event_study_p_values = {}
 
             for e in rel_periods:
-                original_effect = event_study_info[e]['effect']
-                boot_dist = bootstrap_event_study[e]
-
-                event_study_ses[e] = float(np.std(boot_dist, ddof=1))
-                event_study_cis[e] = self._compute_percentile_ci(boot_dist, self.alpha)
-                event_study_p_values[e] = self._compute_bootstrap_pvalue(
-                    original_effect, boot_dist
+                se, ci, p_value = self._compute_effect_bootstrap_stats(
+                    event_study_info[e]['effect'], bootstrap_event_study[e]
                 )
+                event_study_ses[e] = se
+                event_study_cis[e] = ci
+                event_study_p_values[e] = p_value
 
-        # Group effect statistics
+        # Compute bootstrap statistics for group effects
         group_effect_ses = None
         group_effect_cis = None
         group_effect_p_values = None
@@ -1657,14 +1650,12 @@ class CallawaySantAnna:
             group_effect_p_values = {}
 
             for g in groups:
-                original_effect = group_agg_info[g]['effect']
-                boot_dist = bootstrap_group[g]
-
-                group_effect_ses[g] = float(np.std(boot_dist, ddof=1))
-                group_effect_cis[g] = self._compute_percentile_ci(boot_dist, self.alpha)
-                group_effect_p_values[g] = self._compute_bootstrap_pvalue(
-                    original_effect, boot_dist
+                se, ci, p_value = self._compute_effect_bootstrap_stats(
+                    group_agg_info[g]['effect'], bootstrap_group[g]
                 )
+                group_effect_ses[g] = se
+                group_effect_cis[g] = ci
+                group_effect_p_values[g] = p_value
 
         return CSBootstrapResults(
             n_bootstrap=self.n_bootstrap,
@@ -1817,6 +1808,35 @@ class CallawaySantAnna:
 
         return float(p_value)
 
+    def _compute_effect_bootstrap_stats(
+        self,
+        original_effect: float,
+        boot_dist: np.ndarray,
+    ) -> Tuple[float, Tuple[float, float], float]:
+        """
+        Compute bootstrap statistics for a single effect.
+
+        Parameters
+        ----------
+        original_effect : float
+            Original point estimate.
+        boot_dist : np.ndarray
+            Bootstrap distribution of the effect.
+
+        Returns
+        -------
+        se : float
+            Bootstrap standard error.
+        ci : Tuple[float, float]
+            Percentile confidence interval.
+        p_value : float
+            Bootstrap p-value.
+        """
+        se = float(np.std(boot_dist, ddof=1))
+        ci = self._compute_percentile_ci(boot_dist, self.alpha)
+        p_value = self._compute_bootstrap_pvalue(original_effect, boot_dist)
+        return se, ci, p_value
+
     def get_params(self) -> Dict[str, Any]:
         """Get estimator parameters (sklearn-compatible)."""
         return {

{diff_diff-1.2.0.dist-info → diff_diff-1.2.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diff-diff
-Version: 1.2.0
+Version: 1.2.1
 Summary: A library for Difference-in-Differences causal inference analysis
 Author: diff-diff contributors
 License-Expression: MIT

{diff_diff-1.2.0.dist-info → diff_diff-1.2.1.dist-info}/RECORD

@@ -1,4 +1,4 @@
-diff_diff/__init__.py,sha256=Phf8CQcV-bH2obbW0W5TQRgeWyq6Z7cLJxGwwZyNj_A,4169
+diff_diff/__init__.py,sha256=qBIk3EAm-AzDDnul7ruuq1O6rLC8sNdfi9GW8F3h_g4,4169
 diff_diff/bacon.py,sha256=AgQOtGUN-gtMh8q6KsGmOiVF4uCCFG2IexbeKS0mBlQ,36819
 diff_diff/diagnostics.py,sha256=1yOKfauW9RP7alXYlaR8oq5tmol5T24idsfqzyAxBPQ,29572
 diff_diff/estimators.py,sha256=4Jj9xAnF39EWElweFoYUa91YnyELelT8mRDzZKZWj4E,35807
@@ -7,13 +7,13 @@ diff_diff/power.py,sha256=cpdbWG-lxqAKjeAoM_8Un8gYaDwADIeYdWUIkEBfXZg,42800
 diff_diff/prep.py,sha256=bTPXTWajBzdNVPcLQ_IBYjww63DhgS3Z37V9h5tsIgc,46985
 diff_diff/pretrends.py,sha256=NeYjTxC9s_iIj-3beYupiDAZROYaLjFIzbm-76XhqKk,36538
 diff_diff/results.py,sha256=qmostsUy7uqX2VjAT7qSUPS3W1q_K8bqBB2JuUri_-k,22097
-diff_diff/staggered.py,sha256=F-7CPXGtIEhpcaxQnJYXTfz3AvCoqVFOdZsHL7Izy8c,69010
+diff_diff/staggered.py,sha256=DPx6VRlF67IryT8cXYq_gfdclQZEDHIYzQakR3OSvWA,69465
 diff_diff/sun_abraham.py,sha256=ux6igLILW7Y1bh3TK33P781ATUWJS4fkwrjnpTARPlU,41685
 diff_diff/synthetic_did.py,sha256=g_Es1H-Yi1UozGXyoW2ZDK8x7BrEEf73DWNqJD9mpGk,19303
 diff_diff/twfe.py,sha256=iB-hvmOQd97B5_3hWvoaYGCSfcwJbXfg6towhRYD_Bw,11931
 diff_diff/utils.py,sha256=F0vRZ32k065VsP4YcAcUdaF_dom834QOgYj3ZaGS2pI,44211
 diff_diff/visualization.py,sha256=1X074fUk_fZleAlHdSGYc-ihdShY9FJ6WoKmBFCV6oI,51537
-diff_diff-1.2.0.dist-info/METADATA,sha256=hmqUTkQVtv9oPW05KtB87lvl3NkOX4gjc8nhrWOK9qo,71789
-diff_diff-1.2.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-diff_diff-1.2.0.dist-info/top_level.txt,sha256=-7mAFgjEQIA2okDLHlh5pDwBQXnsO1Z85qvtHjZILoQ,10
-diff_diff-1.2.0.dist-info/RECORD,,
+diff_diff-1.2.1.dist-info/METADATA,sha256=-UoafGfh9HkQo_fGGcoSprBm-51fY1E4W7vab_tOPLs,71789
+diff_diff-1.2.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+diff_diff-1.2.1.dist-info/top_level.txt,sha256=-7mAFgjEQIA2okDLHlh5pDwBQXnsO1Z85qvtHjZILoQ,10
+diff_diff-1.2.1.dist-info/RECORD,,