Merge pull request #647 from bashtage/pandas-3-fixes

MAINT: Fixes for pandas and flake

Author: bashtage

linearmodels/system/model.py

@@ -102,7 +102,7 @@ def _missing_weights(
     missing = [key for key in weights if weights[key] is None]
     if missing:
         msg = "Weights not found for equation labels:\n{}".format(", ".join(missing))
-        warnings.warn(msg, UserWarning)
+        warnings.warn(msg, UserWarning, stacklevel=2)
 
 
 def _parameters_from_xprod(
@@ -1917,6 +1917,7 @@ def __init__(
                 "matrix not unadjusted (homoskedastic).  sigma will "
                 "be ignored.",
                 UserWarning,
+                stacklevel=1,
             )
         weight_type = COV_TYPES[weight_type]
         self._weight_est = GMM_W_EST[weight_type](**weight_config)

linearmodels/tests/panel/_utility.py

@@ -200,13 +200,23 @@ def generate_data(
 
 
 def assert_results_equal(res1, res2, test_fit=True, test_df=True, strict=True):
+    from pandas import Index
+
+    def fix_index(x: Series | DataFrame, n: int):
+        if isinstance(x, Series):
+            out = x.iloc[:n].copy()
+        else:
+            out = x.iloc[:n, :n].copy()
+        out.index = Index(out.index.to_list())
+        return out
+
     n = min(res1.params.shape[0], res2.params.shape[0])
 
-    assert_series_equal(res1.params.iloc[:n], res2.params.iloc[:n])
-    assert_series_equal(res1.pvalues.iloc[:n], res2.pvalues.iloc[:n])
-    assert_series_equal(res1.tstats.iloc[:n], res2.tstats.iloc[:n])
-    assert_frame_equal(res1.cov.iloc[:n, :n], res2.cov.iloc[:n, :n])
-    assert_frame_equal(res1.conf_int().iloc[:n], res2.conf_int().iloc[:n])
+    assert_series_equal(res1.params.iloc[:n], fix_index(res2.params, n))
+    assert_series_equal(res1.pvalues.iloc[:n], fix_index(res2.pvalues, n))
+    assert_series_equal(res1.tstats.iloc[:n], fix_index(res2.tstats, n))
+    assert_frame_equal(res1.cov.iloc[:n, :n], fix_index(res2.cov, n))
+    assert_frame_equal(res1.conf_int().iloc[:n], fix_index(res2.conf_int(), n))
 
     assert_allclose(res1.s2, res2.s2)
 

linearmodels/tests/panel/test_panel_ols.py

@@ -286,7 +286,9 @@ def test_panel_entity_lsdv(data):
         d_demean = d.values
 
     xd = np.c_[x.values, d_demean]
-    xd = pd.DataFrame(xd, index=x.index, columns=list(x.columns) + list(d.columns))
+    xd = pd.DataFrame(
+        xd, index=x.index, columns=list(x.columns) + [f"fe_{col}" for col in d.columns]
+    )
 
     ols_mod = IV2SLS(y, xd, None, None)
     res2 = ols_mod.fit(cov_type="unadjusted", debiased=False)
@@ -388,7 +390,9 @@ def test_panel_time_lsdv(large_data):
         d = d - z @ lstsq(z, d, rcond=None)[0]
 
     xd = np.c_[x.values, d]
-    xd = pd.DataFrame(xd, index=x.index, columns=list(x.columns) + d_cols)
+    xd = pd.DataFrame(
+        xd, index=x.index, columns=list(x.columns) + [f"fe_{col}" for col in d_cols]
+    )
 
     ols_mod = IV2SLS(y, xd, None, None)
     res2 = ols_mod.fit(cov_type="unadjusted")
@@ -490,7 +494,11 @@ def test_panel_both_lsdv(data):
 
     xd = np.c_[x.values, d]
     xd = pd.DataFrame(
-        xd, index=x.index, columns=list(x.columns) + list(d1.columns) + list(d2.columns)
+        xd,
+        index=x.index,
+        columns=list(x.columns)
+        + [f"fe_{col}" for col in d1.columns]
+        + [f"te_{col}" for col in d2.columns],
     )
 
     ols_mod = IV2SLS(y, xd, None, None)
@@ -597,7 +605,9 @@ def test_panel_entity_lsdv_weighted(data):
         d = d - z @ lstsq(wz, wd, rcond=None)[0]
 
     xd = np.c_[x.values, d]
-    xd = pd.DataFrame(xd, index=x.index, columns=list(x.columns) + list(d_cols))
+    xd = pd.DataFrame(
+        xd, index=x.index, columns=list(x.columns) + [f"fe_{col}" for col in d_cols]
+    )
 
     ols_mod = IV2SLS(y, xd, None, None, weights=w)
     res2 = ols_mod.fit(cov_type="unadjusted")
@@ -677,7 +687,9 @@ def test_panel_time_lsdv_weighted(large_data):
         d = d - z @ lstsq(wz, wd, rcond=None)[0]
 
     xd = np.c_[x.values, d]
-    xd = pd.DataFrame(xd, index=x.index, columns=list(x.columns) + list(d_cols))
+    xd = pd.DataFrame(
+        xd, index=x.index, columns=list(x.columns) + [f"te_{col}" for col in d_cols]
+    )
 
     ols_mod = IV2SLS(y, xd, None, None, weights=w)
     res2 = ols_mod.fit(cov_type="unadjusted")
@@ -760,7 +772,11 @@ def test_panel_both_lsdv_weighted(data):
 
     xd = np.c_[x.values, d]
     xd = pd.DataFrame(
-        xd, index=x.index, columns=list(x.columns) + list(d1.columns) + list(d2.columns)
+        xd,
+        index=x.index,
+        columns=list(x.columns)
+        + [f"fe_{col}" for col in d1.columns]
+        + [f"te_{col}" for col in d2.columns],
     )
 
     ols_mod = IV2SLS(y, xd, None, None, weights=w)