Multiple Functions require lists of Feols/Feiv/Fepois object; reject FixestMulti #693

pyfixest/estimation/FixestMulti_.py

@@ -386,7 +386,7 @@ def _estimate_all_models(
 
                     self.all_fitted_models[FIT._model_name] = FIT
 
-    def to_list(self):
+    def to_list(self) -> list[Union[Feols, Fepois, Feiv]]:
         """
         Return a list of all fitted models.
 

pyfixest/estimation/feols_.py

@@ -1696,10 +1696,12 @@ def get_performance(self) -> None:
         ssy = np.sum((_Y - np.mean(_Y)) ** 2)
 
         if _has_weights:
+            ssy = np.sum(_Y**2) - np.sum(_Y**2) / np.sum(_weights)
             self._rmse = np.nan
             self._r2 = np.nan
             self._adj_r2 = np.nan
         else:
+            ssy = np.sum((_Y - np.mean(_Y)) ** 2)
             self._rmse = np.sqrt(ssu / _N)
             self._r2 = 1 - (ssu / ssy)
             self._adj_r2 = 1 - (ssu / ssy) * _adj_factor

pyfixest/estimation/multcomp.py

@@ -4,12 +4,16 @@
 import numpy as np
 import pandas as pd
 
+from pyfixest.estimation.feiv_ import Feiv
 from pyfixest.estimation.feols_ import Feols
 from pyfixest.estimation.fepois_ import Fepois
+from pyfixest.estimation.FixestMulti_ import FixestMulti
 from pyfixest.report.summarize import _post_processing_input_checks
 
+ModelInputType = Union[FixestMulti, list[Union[Feols, Fepois, Feiv]]]
 
-def bonferroni(models: list[Union[Feols, Fepois]], param: str) -> pd.DataFrame:
+
+def bonferroni(models: ModelInputType, param: str) -> pd.DataFrame:
     """
     Compute Bonferroni adjusted p-values for multiple hypothesis testing.
 
@@ -18,9 +22,8 @@ def bonferroni(models: list[Union[Feols, Fepois]], param: str) -> pd.DataFrame:
 
     Parameters
     ----------
-    models : list[Feols, Fepois], Feols or Fepois
-        A list of models for which the p-values should be adjusted, or a Feols or
-        Fepois object.
+    models : A supported model object (Feols, Fepois, Feiv, FixestMulti) or a list of
+            Feols, Fepois & Feiv models.
     param : str
         The parameter for which the p-values should be adjusted.
 
@@ -65,7 +68,7 @@ def bonferroni(models: list[Union[Feols, Fepois]], param: str) -> pd.DataFrame:
 
 
 def rwolf(
-    models: list[Union[Feols, Fepois]],
+    models: Union[FixestMulti, list[Union[Feols, Fepois]]],
     param: str,
     reps: int,
     seed: int,

pyfixest/report/summarize.py

@@ -1,4 +1,5 @@
 import re
+from collections.abc import ValuesView
 from typing import Optional, Union
 
 import numpy as np
@@ -13,9 +14,13 @@
 from pyfixest.report.utils import _relabel_expvar
 from pyfixest.utils.dev_utils import _select_order_coefs
 
+ModelInputType = Union[
+    FixestMulti, Feols, Fepois, Feiv, list[Union[Feols, Fepois, Feiv]]
+]
+
 
 def etable(
-    models: Union[list[Union[Feols, Fepois, Feiv]], FixestMulti],
+    models: ModelInputType,
     type: str = "gt",
     signif_code: Optional[list] = None,
     coef_fmt: str = "b \n (se)",
@@ -40,8 +45,8 @@
 
     Parameters
     ----------
-    models : list
-        A list of models of type Feols, Feiv, Fepois.
+    models : A supported model object (Feols, Fepois, Feiv, FixestMulti) or a list of
+            Feols, Fepois & Feiv models.
     type : str, optional
         Type of output. Either "df" for pandas DataFrame, "md" for markdown,
         "gt" for great_tables, or "tex" for LaTeX table. Default is "gt".
@@ -132,11 +137,6 @@
             signif_code[0] < signif_code[1] < signif_code[2]
         ), "signif_code must be in increasing order"
 
-    # Check if models is of type FixestMulti
-    # If so, convert it to a list of models
-    if isinstance(models, FixestMulti):
-        models = models.to_list()
-
     models = _post_processing_input_checks(models)
 
     if custom_stats is None:
@@ -464,9 +464,7 @@
     return None
 
 
-def summary(
-    models: Union[list[Union[Feols, Fepois, Feiv]], FixestMulti], digits: int = 3
-) -> None:
+def summary(models: ModelInputType, digits: int = 3) -> None:
     """
     Print a summary of estimation results for each estimated model.
 
@@ -476,8 +474,8 @@
 
     Parameters
     ----------
-    models : list[Union[Feols, Fepois, Feiv]] or FixestMulti.
-            The models to be summarized.
+    models : A supported model object (Feols, Fepois, Feiv, FixestMulti) or a list of
+            Feols, Fepois & Feiv models.
     digits : int, optional
         The number of decimal places to round the summary statistics to. Default is 3.
 
@@ -550,8 +548,8 @@
 
 
 def _post_processing_input_checks(
-    models: Union[list[Union[Feols, Fepois, Feiv]], FixestMulti],
-) -> list[Union[Feols, Fepois]]:
+    models: ModelInputType,
+) -> list[Union[Feols, Fepois, Feiv]]:
     """
     Perform input checks for post-processing models.
 
@@ -572,34 +570,23 @@
         TypeError: If the models argument is not of the expected type.
 
     """
-    # check if models instance of Feols or Fepois
-    if isinstance(models, (Feols, Fepois)):
-        models = [models]
-
-    else:
-        if isinstance(models, (list, type({}.values()))):
-            for model in models:
-                if not isinstance(model, (Feols, Fepois)):
-                    raise TypeError(
-                        f"""
-                        Each element of the passed list needs to be of type Feols
-                        or Fepois, but {type(model)} was passed. If you want to
-                        summarize a FixestMulti object, please use FixestMulti.to_list()
-                        to convert it to a list of Feols or Fepois instances.
-                        """
-                    )
-
+    models_list: list[Union[Feols, Fepois, Feiv]] = []
+
+    if isinstance(models, (Feols, Fepois, Feiv)):
+        models_list = [models]
+    elif isinstance(models, FixestMulti):
+        models_list = models.to_list()
+    elif isinstance(models, (list, ValuesView)):
+        if all(isinstance(m, (Feols, Fepois, Feiv)) for m in models):
+            models_list = models
         else:
             raise TypeError(
-                """
-                The models argument must be either a list of Feols or Fepois instances, or
-                simply a single Feols or Fepois instance. The models argument does not accept instances
-                of type FixestMulti - please use models.to_list() to convert the FixestMulti
-                instance to a list of Feols or Fepois instances.
-                """
+                "All elements in the models list must be instances of Feols, Feiv, or Fepois."
             )
+    else:
+        raise TypeError("Invalid type for models argument.")
 
-    return models
+    return models_list
 
 
 def _tabulate_etable_md(df, n_coef, n_fixef, n_models, n_model_stats):

pyfixest/report/visualize.py

@@ -23,10 +23,16 @@
 from pyfixest.estimation.feiv_ import Feiv
 from pyfixest.estimation.feols_ import Feols
 from pyfixest.estimation.fepois_ import Fepois
+from pyfixest.estimation.FixestMulti_ import FixestMulti
 from pyfixest.report.summarize import _post_processing_input_checks
 from pyfixest.report.utils import _relabel_expvar
 from pyfixest.utils.dev_utils import _select_order_coefs
 
+ModelInputType = Union[
+    FixestMulti, Feols, Fepois, Feiv, list[Union[Feols, Fepois, Feiv]]
+]
+
+
 LetsPlot.setup_html()
 
 
@@ -60,7 +66,7 @@ def set_figsize(
 
 
 def iplot(
-    models,
+    models: ModelInputType,
     alpha: float = 0.05,
     figsize: Optional[tuple[int, int]] = None,
     yintercept: Union[int, str, None] = None,
@@ -83,9 +89,8 @@ def iplot(
 
     Parameters
     ----------
-    models : list or object
-        A list of fitted models of type `Feols` or
-        `Fepois`, or just a single model.
+    models : A supported model object (Feols, Fepois, Feiv, FixestMulti) or a list of
+            Feols, Fepois & Feiv models.
     figsize : tuple or None, optional
         The size of the figure. If None, the default size is used.
     alpha : float
@@ -157,8 +162,8 @@ def iplot(
             "The 'joint' parameter is only available for a single model, i.e. objects of type FixestMulti are not supported."
         )
 
-    df_all = []
-    all_icovars = []
+    df_all: list[pd.DataFrame] = []
+    all_icovars: list[str] = []
 
     if keep is None:
         keep = []
@@ -205,7 +210,7 @@ def iplot(
 
 
 def coefplot(
-    models: list,
+    models: ModelInputType,
     alpha: float = 0.05,
     figsize: Optional[tuple[int, int]] = None,
     yintercept: float = 0,
@@ -227,8 +232,8 @@ def coefplot(
 
     Parameters
     ----------
-    models : list or object
-        A list of fitted models of type `Feols` or `Fepois`, or just a single model.
+    models : A supported model object (Feols, Fepois, Feiv, FixestMulti) or a list of
+            Feols, Fepois & Feiv models.
     figsize : tuple or None, optional
         The size of the figure. If None, the default size is used.
     alpha : float

tests/test_api.py

@@ -13,11 +13,17 @@ def test_api():
     fit2 = pf.estimation.fepois(
         "Y ~ X1 + X2 + f2 | f1", data=df2, vcov={"CRV1": "f1+f2"}
     )
+    fit_multi = pf.feols("Y + Y2 ~ X1", data=df2)
+
     pf.summary(fit1)
     pf.report.summary(fit2)
     pf.etable([fit1, fit2])
     pf.coefplot([fit1, fit2])
 
+    pf.summary(fit_multi)
+    pf.etable(fit_multi)
+    pf.coefplot(fit_multi)
+
 
 def test_feols_args():
     """

tests/test_multcomp.py

@@ -5,6 +5,7 @@
 from rpy2.robjects import pandas2ri
 from rpy2.robjects.packages import importr
 
+import pyfixest as pf
 from pyfixest.estimation.estimation import feols
 from pyfixest.estimation.multcomp import _get_rwolf_pval, bonferroni, rwolf
 from pyfixest.utils.set_rpy2_path import update_r_paths
@@ -195,3 +196,20 @@ def test_sampling_scheme(seed, reps):
     assert (
         np.abs(percent_diff) < 1.0
     ), f"Percentage difference is too large: {percent_diff}%"
+
+
+@pytest.mark.extended
+def test_multi_vs_list(seeds, reps):
+    "Test that lists of models and FixestMulti input produce identical results."
+    seed = 1232
+    reps = 100
+    data = pf.get_data(N=100)
+
+    fit_all = pf.feols("Y + Y2 ~ X1 + X2", data=data)
+    fit1 = pf.feols("Y ~ X1 + X2", data=data)
+    fit2 = pf.feols("Y2 ~ X1 + X2", data=data)
+
+    assert bonferroni(fit_all, "X1").equals(bonferroni([fit1, fit2], "X1"))
+    assert rwolf(fit_all, "X1", seed=seed, reps=reps).equals(
+        rwolf([fit1, fit2], "X1", seed=seed, reps=reps)
+    )