Merge pull request #296 from flatironinstitute/hotfix_repr

Hotfix repr

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "nemos"
-version = "0.2.0"
+version = "0.2.1"
 authors = [{name = "nemos authors"}]
 description = "NEural MOdelS, a statistical modeling framework for neuroscience."
 readme = "README.md"

src/nemos/__init__.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-__version__ = "0.2.0"
+__version__ = "0.2.1"
 
 from . import (
     basis,

src/nemos/glm.py

@@ -1126,6 +1126,12 @@ def _get_optimal_solver_params_config(self):
     def __repr__(self):
         return format_repr(self, multiline=True)
 
+    def __sklearn_clone__(self) -> GLM:
+        """Clone the PopulationGLM, dropping feature_mask"""
+        params = self.get_params(deep=False)
+        klass = self.__class__(**params)
+        return klass
+
 
 class PopulationGLM(GLM):
     """
@@ -1633,7 +1639,7 @@ def _predict(
             + bs
         )
 
-    def __sklearn_clone__(self) -> GLM:
+    def __sklearn_clone__(self) -> PopulationGLM:
         """Clone the PopulationGLM, dropping feature_mask"""
         params = self.get_params(deep=False)
         params.pop("feature_mask")

src/nemos/utils.py

@@ -534,7 +534,7 @@ def format_repr(
         )
         if repr_param:
             if k in use_name_keys:
-                v = v.__name__
+                v = getattr(v, "__name__", repr(v))
             elif isinstance(v, str):
                 v = repr(v)
             disp_params.append(f"{k}={v}")

tests/test_glm.py

@@ -1478,12 +1478,16 @@ def test_deviance_against_statsmodels(self, poissonGLM_model_instantiation):
     def test_compatibility_with_sklearn_cv(self, poissonGLM_model_instantiation):
         X, y, model, true_params, firing_rate = poissonGLM_model_instantiation
         param_grid = {"solver_name": ["BFGS", "GradientDescent"]}
-        GridSearchCV(model, param_grid).fit(X, y)
+        cls = GridSearchCV(model, param_grid).fit(X, y)
+        # check that the repr works after cloning
+        repr(cls)
 
     def test_compatibility_with_sklearn_cv_gamma(self, gammaGLM_model_instantiation):
         X, y, model, true_params, firing_rate = gammaGLM_model_instantiation
         param_grid = {"solver_name": ["BFGS", "GradientDescent"]}
-        GridSearchCV(model, param_grid).fit(X, y)
+        cls = GridSearchCV(model, param_grid).fit(X, y)
+        # check that the repr works after cloning
+        repr(cls)
 
     @pytest.mark.parametrize(
         "regr_setup, glm_class",
@@ -3572,12 +3576,16 @@ def test_deviance_against_statsmodels(self, poisson_population_GLM_model):
     def test_compatibility_with_sklearn_cv(self, poisson_population_GLM_model):
         X, y, model, true_params, firing_rate = poisson_population_GLM_model
         param_grid = {"solver_name": ["BFGS", "GradientDescent"]}
-        GridSearchCV(model, param_grid).fit(X, y)
+        cls = GridSearchCV(model, param_grid).fit(X, y)
+        # check that the repr works after cloning
+        repr(cls)
 
     def test_compatibility_with_sklearn_cv_gamma(self, gamma_population_GLM_model):
         X, y, model, true_params, firing_rate = gamma_population_GLM_model
         param_grid = {"solver_name": ["BFGS", "GradientDescent"]}
-        GridSearchCV(model, param_grid).fit(X, y)
+        cls = GridSearchCV(model, param_grid).fit(X, y)
+        # check that the repr works after cloning
+        repr(cls)
 
     def test_sklearn_clone(self, poisson_population_GLM_model):
         X, y, model, true_params, firing_rate = poisson_population_GLM_model

tests/test_utils.py

@@ -1,5 +1,6 @@
 import warnings
 from contextlib import nullcontext as does_not_raise
+from copy import deepcopy
 
 import jax
 import jax.numpy as jnp
@@ -595,6 +596,8 @@ def __repr__(self):
         (Example(a=0, b=False, c=None), None, [], "Example(a=0, b=False, d=1)"),
         # Falsey values excluded2
         (Example(a=0, b=[], c={}), None, [], "Example(a=0, d=1)"),
+        # function without the __name__
+        (nmo.observation_models.PoissonObservations(deepcopy(jax.numpy.exp)),None, [], "PoissonObservations(inverse_link_function=<PjitFunction>)")
     ],
 )
 def test_format_repr(obj, exclude_keys, use_name_keys, expected):