wavetrainer 0.1.4__tar.gz → 0.1.6__tar.gz

{wavetrainer-0.1.4/wavetrainer.egg-info → wavetrainer-0.1.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: wavetrainer
-Version: 0.1.4
+Version: 0.1.6
 Summary: A library for automatically finding the optimal model within feature and hyperparameter space.
 Home-page: https://github.com/8W9aG/wavetrainer
 Author: Will Sackfield

{wavetrainer-0.1.4 → wavetrainer-0.1.6}/setup.py

@@ -23,7 +23,7 @@ def install_requires() -> typing.List[str]:
 
 setup(
     name='wavetrainer',
-    version='0.1.4',
+    version='0.1.6',
     description='A library for automatically finding the optimal model within feature and hyperparameter space.',
     long_description=long_description,
     long_description_content_type='text/markdown',

{wavetrainer-0.1.4 → wavetrainer-0.1.6}/wavetrainer/__init__.py

@@ -2,5 +2,5 @@
 
 from .create import create
 
-__VERSION__ = "0.1.4"
+__VERSION__ = "0.1.6"
 __all__ = ("create",)

{wavetrainer-0.1.4 → wavetrainer-0.1.6}/wavetrainer/model/catboost/catboost_model.py

@@ -122,9 +122,10 @@ class CatboostModel(Model):
         )
         self._early_stopping_rounds = trial.suggest_int(_EARLY_STOPPING_ROUNDS, 10, 500)
         self._best_iteration = trial.user_attrs.get(_BEST_ITERATION_KEY)
-        loss_function = trial.suggest_categorical(
-            _LOSS_FUNCTION_KEY, [_DEFAULT_LOSS_FUNCTION, _FOCALLOSS_LOSS_FUNCTION]
-        )
+        loss_functions = [_DEFAULT_LOSS_FUNCTION]
+        if torch.cuda.is_available():
+            loss_functions.append(_FOCALLOSS_LOSS_FUNCTION)
+        loss_function = trial.suggest_categorical(_LOSS_FUNCTION_KEY, loss_functions)
         self._loss_function = loss_function
         if loss_function == _FOCALLOSS_LOSS_FUNCTION:
             self._gamma = trial.suggest_float(_GAMMA_KEY, 0.5, 5.0)

{wavetrainer-0.1.4 → wavetrainer-0.1.6}/wavetrainer/model/model_router.py

@@ -170,8 +170,7 @@ class ModelRouter(Model):
                     ),
                     axis=1,
                 )
-            w += pred["__wavetrain_error_weight"]
-            w = w.clip(lower=0.0)
+            w += pred["__wavetrain_error_weight"].fillna(0.0).clip(lower=0.000001)
 
         return self
 

{wavetrainer-0.1.4 → wavetrainer-0.1.6}/wavetrainer/reducer/pca_reducer.py

@@ -63,7 +63,7 @@ class PCAReducer(Reducer):
         if self._embedding_cols is None:
             return self
         for k, v in self._pcas.items():
-            v.fit(df[self._embedding_dict[k]])
+            v.fit(df[self._embedding_dict[k]].dropna())
         return self
 
     def transform(self, df: pd.DataFrame) -> pd.DataFrame:
@@ -71,7 +71,7 @@ class PCAReducer(Reducer):
             return df
         for k, v in self._pcas.items():
             cols = self._embedding_dict[k]
-            compressed_embedding = v.transform(df[cols])
+            compressed_embedding = v.transform(df[cols].fillna(0.0))
             embedding_len = compressed_embedding.shape[0]
             df[cols[:embedding_len]] = compressed_embedding
             df = df.drop(columns=cols[embedding_len:])

{wavetrainer-0.1.4 → wavetrainer-0.1.6/wavetrainer.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: wavetrainer
-Version: 0.1.4
+Version: 0.1.6
 Summary: A library for automatically finding the optimal model within feature and hyperparameter space.
 Home-page: https://github.com/8W9aG/wavetrainer
 Author: Will Sackfield