fix wrong change

mlinsights/helpers/pipeline.py

@@ -120,7 +120,7 @@ def to_str(self, nrows=5):
         """
         rows = [f"BaseEstimatorDebugInformation({self.model.__class__.__name__})"]
         for k in sorted(self.inputs):
-            assert k in self.outputs, f"Unable to find output for method '{k}'."
+            assert k in self.outputs, f"Unable to find output for method {k!r}."
             rows.append("  " + k + "(")
             self.display(self.inputs[k], nrows)
             rows.append(textwrap.indent(self.display(self.inputs[k], nrows), "   "))

mlinsights/mlbatch/cache_model.py

@@ -80,7 +80,7 @@ def as_key(params):
             elif v is None:
                 sv = ""
             else:
-                raise TypeError(f"Unable to create a key with value '{k}':{v}")
+                raise TypeError(f"Unable to create a key with value {k!r}:{v!r}")
             els.append((k, sv))
         return str(els)
 

mlinsights/mlmodel/categories_to_integers.py

@@ -70,10 +70,11 @@ def fit(self, X, y=None, **fit_params):
         """
         if not isinstance(X, pandas.DataFrame):
             raise TypeError(f"this transformer only accept Dataframes, not {type(X)}")
-        if self.columns:
-            columns = self.columns
-        else:
-            columns = [c for c, d in zip(X.columns, X.dtypes) if d in (object,)]
+        columns = (
+            self.columns
+            if self.columns
+            else [c for c, d in zip(X.columns, X.dtypes) if d in (object, str)]
+        )
 
         self._fit_columns = columns
         max_cat = max(len(X) // 2 + 1, 10000)
@@ -86,7 +87,7 @@ def fit(self, X, y=None, **fit_params):
                 raise ValueError(
                     f"Too many categories ({nb}) for one column '{c}' max_cat={max_cat}"
                 )
-            self._categories[c] = dict(enumerate(list(sorted(distinct))))
+            self._categories[c] = {c: i for i, c in enumerate(list(sorted(distinct)))}
 
         self._schema = self._build_schema()
         return self
@@ -181,7 +182,7 @@ def transform(v, vec):
                                 lv.append("...")
                             m = "\n".join(map(str, lv))
                             raise ValueError(
-                                f"Unable to find category value {k}: {v} "
+                                f"Unable to find category value {k!r}: {v!r} "
                                 f"type(v)={type(v)} among\n{m}"
                             )
                         p = pos[k]

mlinsights/mlmodel/classification_kmeans.py

@@ -147,7 +147,7 @@ def set_params(self, **values):
             elif k.startswith("c_"):
                 pc[k[2:]] = v
             else:
-                raise ValueError(f"Unexpected parameter name '{k}'")
+                raise ValueError(f"Unexpected parameter name {k!r}")
         self.clus.set_params(**pc)
         self.estimator.set_params(**pe)
 

mlinsights/mlmodel/kmeans_l1.py

@@ -173,7 +173,7 @@ def _validate_center_shape(X, k, centers):
         """Check if centers is compatible with X and n_clusters"""
         assert centers.shape[0] == k, (
             f"The shape of the initial centers {centers.shape} does not "
-            f"match the number of clusters {k}."
+            f"match the number of clusters {k!r}."
         )
         assert centers.shape[1] == X.shape[1], (
             f"The shape of the initial centers {centers.shape} does not "

mlinsights/mlmodel/sklearn_testing.py

@@ -131,7 +131,7 @@ def run_test_sklearn_clone(fct_model, ext=None, copy_fitted=False):
                 ext.assertEqual(p1[k], p2[k])
             except AssertionError as e:
                 raise AssertionError(
-                    f"Difference for key '{k}'\n==1 {p1[k]}\n==2 {p2[k]}"
+                    f"Difference for key {k!r}\n==1 {p1[k]}\n==2 {p2[k]}"
                 ) from e
     return conv, cloned
 
@@ -303,7 +303,7 @@ def adjust(obj1, obj2):
                     v1 = getattr(obj1, k)
                     setattr(obj2, k, clone_with_fitted_parameters(v1))
                 else:
-                    raise RuntimeError(f"Cloned object is missing '{k}' in {obj2}.")
+                    raise RuntimeError(f"Cloned object is missing {k!r} in {obj2}.")
 
     if isinstance(est, BaseEstimator):
         cloned = clone(est)

mlinsights/sklapi/sklearn_base.py

@@ -91,12 +91,12 @@ def compare_params(
         for k in p1:
             if k not in p2:
                 if exc:
-                    raise KeyError(f"Key '{k}' was removed.")
+                    raise KeyError(f"Key {k!r} was removed.")
                 return False
         for k in p2:
             if k not in p1:
                 if exc:
-                    raise KeyError(f"Key '{k}' was added.")
+                    raise KeyError(f"Key {k!r} was added.")
                 return False
         for k in sorted(p1):
             v1, v2 = p1[k], p2[k]
@@ -133,7 +133,7 @@ def compare_params(
             if not b:
                 if exc:
                     raise ValueError(
-                        f"Values for key '{k}' are different.\n---\n{v1}\n---\n{v2}"
+                        f"Values for key {k!r} are different.\n---\n{v1}\n---\n{v2}"
                     )
                 return False
         return True

mlinsights/sklapi/sklearn_base_transform_learner.py

@@ -166,7 +166,7 @@ def set_params(self, **values):
             del values["method"]
         for k in values:
             if not k.startswith("model__"):
-                raise ValueError(f"Parameter '{k}' must start with 'model__'.")
+                raise ValueError(f"Parameter {k!r} must start with 'model__'.")
         d = len("model__")
         pars = {k[d:]: v for k, v in values.items()}
         self.model.set_params(**pars)

mlinsights/sklapi/sklearn_base_transform_stacking.py

@@ -163,7 +163,7 @@ def set_params(self, **values):
             del values["method"]
         for k, _v in values.items():
             if not k.startswith("models_"):
-                raise ValueError(f"Parameter '{k}' must start with 'models_'.")
+                raise ValueError(f"Parameter {k!r} must start with 'models_'.")
         d = len("models_")
         pars = [{} for m in self.models]
         for k, v in values.items():