PyFixest 0.8.10.1: add black workflow (#207)

* add black action * run black * revert notebook
py-econometrics · Nov 5, 2023 · b5d37d7 · b5d37d7
1 parent c4d1e41
commit b5d37d7
Show file tree

Hide file tree

Showing 22 changed files with 561 additions and 464 deletions.
diff --git a/.github/workflows/black.yml b/.github/workflows/black.yml
@@ -0,0 +1,10 @@
+name: Lint
+
+on: [push, pull_request]
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+      - uses: psf/black@stable
diff --git a/docs/tutorial.ipynb b/docs/tutorial.ipynb
@@ -1121,6 +1121,7 @@
    ],
    "source": [
     "from pyfixest.summarize import etable\n",
+    "\n",
     "etable([fit, fit2])"
    ]
   },
@@ -1273,6 +1274,7 @@
    ],
    "source": [
     "from lets_plot import *\n",
+    "\n",
     "LetsPlot.setup_html()\n",
     "\n",
     "multi_fit.coefplot().show()"
@@ -1445,7 +1447,7 @@
     "import numpy as np\n",
     "from pyfixest.experimental.did import did2s\n",
     "\n",
-    "file_path = '../pyfixest/experimental/data/df_het.csv'\n",
+    "file_path = \"../pyfixest/experimental/data/df_het.csv\"\n",
     "df_het = pd.read_csv(file_path)\n",
     "df_het.head()"
    ]
@@ -1555,22 +1557,24 @@
    "source": [
     "fit_did2s = did2s(\n",
     "    df_het,\n",
-    "    yname = \"dep_var\",\n",
-    "    first_stage = \"~ 0 | state + year\",\n",
-    "    second_stage = \"~i(rel_year)\",\n",
-    "    treatment = \"treat\",\n",
-    "    cluster = \"state\",\n",
-    "    i_ref1 = [-1.0, np.inf],\n",
+    "    yname=\"dep_var\",\n",
+    "    first_stage=\"~ 0 | state + year\",\n",
+    "    second_stage=\"~i(rel_year)\",\n",
+    "    treatment=\"treat\",\n",
+    "    cluster=\"state\",\n",
+    "    i_ref1=[-1.0, np.inf],\n",
     ")\n",
     "\n",
     "fit_twfe = feols(\n",
     "    \"dep_var ~ i(rel_year) | state + year\",\n",
     "    df_het,\n",
-    "    i_ref1 = [-1.0, np.inf],\n",
-    "    vcov = {\"CRV1\": \"state\"}\n",
+    "    i_ref1=[-1.0, np.inf],\n",
+    "    vcov={\"CRV1\": \"state\"},\n",
     ")\n",
     "\n",
-    "iplot([fit_did2s, fit_twfe], coord_flip=False, figsize = (900, 400), title = \"TWFE vs DID2S\")"
+    "iplot(\n",
+    "    [fit_did2s, fit_twfe], coord_flip=False, figsize=(900, 400), title=\"TWFE vs DID2S\"\n",
+    ")"
    ]
   },
   {
@@ -1601,21 +1605,21 @@
     "from pyfixest.summarize import etable\n",
     "\n",
     "fit_twfe = event_study(\n",
-    "    data = df_het,\n",
-    "    yname = \"dep_var\",\n",
-    "    idname= \"state\",\n",
-    "    tname = \"year\",\n",
-    "    gname = \"g\",\n",
-    "    estimator = \"twfe\"\n",
+    "    data=df_het,\n",
+    "    yname=\"dep_var\",\n",
+    "    idname=\"state\",\n",
+    "    tname=\"year\",\n",
+    "    gname=\"g\",\n",
+    "    estimator=\"twfe\",\n",
     ")\n",
     "\n",
     "fit_did2s = event_study(\n",
-    "    data = df_het,\n",
-    "    yname = \"dep_var\",\n",
-    "    idname= \"state\",\n",
-    "    tname = \"year\",\n",
-    "    gname = \"g\",\n",
-    "    estimator = \"did2s\"\n",
+    "    data=df_het,\n",
+    "    yname=\"dep_var\",\n",
+    "    idname=\"state\",\n",
+    "    tname=\"year\",\n",
+    "    gname=\"g\",\n",
+    "    estimator=\"did2s\",\n",
     ")\n",
     "\n",
     "etable([fit_twfe, fit_did2s])"

diff --git a/pyfixest/FixestMulti.py b/pyfixest/FixestMulti.py
@@ -182,8 +182,10 @@ def _estimate_all_models(
                         na_index,
                         na_index_str,
                         _icovars,
-                        X_is_empty
-                    ) = model_matrix_fixest(fml=fml, data=_data, i_ref1 = _i_ref1, i_ref2 = _i_ref2)
+                        X_is_empty,
+                    ) = model_matrix_fixest(
+                        fml=fml, data=_data, i_ref1=_i_ref1, i_ref2=_i_ref2
+                    )
 
                     weights = np.ones((Y.shape[0], 1))
 
@@ -250,7 +252,7 @@ def _estimate_all_models(
                                 X=Xd,
                                 weights=weights,
                                 coefnames=coefnames,
-                                collin_tol=collin_tol
+                                collin_tol=collin_tol,
                             )
 
                         # special case: sometimes it is useful to fit models as "Y ~ 0 | f1 + f2" to demean Y and to use the predict() method
@@ -310,7 +312,6 @@ def _estimate_all_models(
                     # FEOLS.split_log = x
 
                     if not FIT._X_is_empty:
-
                         # inference
                         vcov_type = _get_vcov_type(vcov, fval)
                         FIT.vcov(vcov=vcov_type)
@@ -640,7 +641,6 @@ def get_fml(
         str: The formula string for the regression.
     """
 
-
     fml = f"{depvar} ~ {covar}"
 
     if endogvars is not None:

diff --git a/pyfixest/demean.py b/pyfixest/demean.py
@@ -5,7 +5,6 @@
 from typing import Any, Sequence, Optional
 
 
-
 def demean_model(
     Y: pd.DataFrame,
     X: pd.DataFrame,
@@ -39,7 +38,6 @@ def demean_model(
             - Id (pd.DataFrame or None): A DataFrame of the demeaned Instruments. None if no IV.
     """
 
-
     YX = pd.concat([Y, X], axis=1)
 
     yx_names = YX.columns
@@ -91,11 +89,9 @@ def demean_model(
             # not data demeaned yet for NA combination
 
             if drop_singletons:
-
                 dropped_singleton_indices = _detect_singletons(fe)
 
                 if np.any(dropped_singleton_indices == True):
-
                     print(
                         np.sum(dropped_singleton_indices),
                         "observations are dropped due to singleton fixed effects.",
@@ -107,7 +103,7 @@ def demean_model(
 
             weights = np.ones(YX.shape[0])
 
-            YX_demeaned, success = demean(x = YX, flist = fe, weights = weights)
+            YX_demeaned, success = demean(x=YX, flist=fe, weights=weights)
             if success == False:
                 raise ValueError("Demeaning failed after 100_000 iterations.")
 
@@ -176,7 +172,7 @@ def demean(
     x: np.ndarray,
     flist: np.ndarray,
     weights: np.ndarray,
-    tol: float = 1e-08,             # note: fixest uses 1e-06, but potentially different tolerance criterion
+    tol: float = 1e-08,  # note: fixest uses 1e-06, but potentially different tolerance criterion
     maxiter: int = 100_000,
 ) -> Tuple[np.ndarray, bool]:
     n_samples, n_features = x.shape
@@ -228,9 +224,8 @@ def demean(
     return (res, success)
 
 
-#@nb.jit(nopython=False)
+# @nb.jit(nopython=False)
 def _detect_singletons(ids):
-
     """
     Detect singleton fixed effects
     Args:have a
@@ -249,7 +244,6 @@ def _detect_singletons(ids):
 
     while True:
         for x in range(k):
-
             col = ids[:, x]
             col_tmp = ids_tmp[:, x]
 
@@ -263,18 +257,14 @@ def _detect_singletons(ids):
                 singleton_idx[np.isin(col, idx)] = True
                 singleton_idx_tmp[np.isin(col_tmp, idx)] = True
 
-                ids_tmp = ids_tmp[~singleton_idx_tmp,:].astype(ids_tmp.dtype)
+                ids_tmp = ids_tmp[~singleton_idx_tmp, :].astype(ids_tmp.dtype)
                 singleton_idx_tmp = singleton_idx_tmp[~singleton_idx_tmp]
 
             break
 
-
         if np.array_equal(singleton_idx_tmp, singleton_idx_tmp_old):
             break
 
         singleton_idx_tmp_old = singleton_idx_tmp.copy()
 
-
-
-
     return singleton_idx
diff --git a/pyfixest/estimation.py b/pyfixest/estimation.py
@@ -14,7 +14,7 @@ def feols(
     fixef_rm: str = "none",
     collin_tol: float = 1e-10,
     i_ref1: Optional[Union[list, str]] = None,
-    i_ref2: Optional[Union[list, str]] = None
+    i_ref2: Optional[Union[list, str]] = None,
 ) -> Union[Feols, FixestMulti]:
     """
 
@@ -156,7 +156,7 @@ def fepois(
     iwls_maxiter: int = 25,
     collin_tol: float = 1e-10,
     i_ref1: Optional[Union[list, str]] = None,
-    i_ref2: Optional[Union[list, str]] = None
+    i_ref2: Optional[Union[list, str]] = None,
 ) -> Union[Fepois, FixestMulti]:
     """
     # fepois
@@ -303,8 +303,12 @@ def _estimation_input_checks(fml, data, vcov, ssc, fixef_rm, collin_tol, i_ref1)
     if not collin_tol < 1:
         raise ValueError("collin_tol must be less than one")
 
-    assert i_ref1 is None or isinstance(i_ref1, (list, str, int, bool, float)), "i_ref1 must be either None, a list, string, int, bool, or float"
+    assert i_ref1 is None or isinstance(
+        i_ref1, (list, str, int, bool, float)
+    ), "i_ref1 must be either None, a list, string, int, bool, or float"
     # check that if i_ref1 is a list, all elements are of the same type
     if isinstance(i_ref1, list):
         assert len(i_ref1) > 0, "i_ref1 must not be an empty list"
-        assert all(isinstance(x, type(i_ref1[0])) for x in i_ref1), "i_ref1 must be a list of elements of the same type"
+        assert all(
+            isinstance(x, type(i_ref1[0])) for x in i_ref1
+        ), "i_ref1 must be a list of elements of the same type"
diff --git a/pyfixest/exceptions.py b/pyfixest/exceptions.py
@@ -53,8 +53,10 @@ class NonConvergenceError(Exception):
 class MatrixNotFullRankError(Exception):
     pass
 
+
 class EmptyDesignMatrixError(Exception):
     pass
 
+
 class InvalidReferenceLevelError(Exception):
-    pass
+    pass
diff --git a/pyfixest/experimental/__init__.py b/pyfixest/experimental/__init__.py
@@ -1,4 +1,6 @@
-print("""
+print(
+    """
       You have loaded the 'pyfixest.experimental' module. Code in this module is not yet stable and
       may change in the future. It may also not be sufficiently unit tested (if at all) or documented.
-      """)
+      """
+)