add test for stacked models

Author: kic

pandas-ml-utils-torch/pandas_ml_utils_torch_test/test_pytorch_model.py

@@ -7,7 +7,7 @@
 from pandas_ml_common import naive_splitter
 from pandas_ml_utils import pd, FeaturesAndLabels
 from pandas_ml_utils.ml.model.base_model import AutoEncoderModel
-from pandas_ml_utils_test.ml.data.model.test_abstract_model import TestAbstractModel
+from pandas_ml_utils_test.ml.model.test_abstract_model import TestAbstractModel
 from pandas_ml_utils_torch import PytorchModel, PytorchAutoEncoderModel, PytorchNN
 
 

pandas-ml-utils/pandas_ml_utils/df_patching/model_patch.py

@@ -113,7 +113,11 @@ def fit(self,
         trails = None
         model = model_provider()
         kwargs = merge_kwargs(model.features_and_labels.kwargs, model.kwargs, kwargs)
-        typemap_fitting = {SubModelFeature: lambda df, model, **kwargs: model.fit(df, **kwargs), **self._type_mapping}
+
+        def fit_submodel(df, model, **kwargs):
+            return model.fit(df, **kwargs)
+
+        typemap_fitting = {SubModelFeature: fit_submodel, **self._type_mapping}
         frames: FeaturesWithLabels = model.features_and_labels(
             df, extract_feature_labels_weights, type_map=typemap_fitting,
             splitter=splitter, filter=filter, cross_validation=cross_validation, epochs=epochs, batch_size=batch_size,

pandas-ml-utils/pandas_ml_utils/ml/model/base_model.py

@@ -265,9 +265,10 @@ def __init__(self, name: str, model: Model):
     def fit(self, df: Typing.PatchedDataFrame, **kwargs):
         _log.info(f"fitting submodel: {self.name}")
         with df.model() as m:
-            fit = m.fit(df, **kwargs)
+            fit = m.fit(self.model, **kwargs)
             self.model = fit.model
 
+        _log.info(f"fitted submodel: {fit}")
         return self.predict(df, **kwargs)
 
     def predict(self, df: Typing.PatchedDataFrame, **kwargs):

pandas-ml-utils/pandas_ml_utils_test/ml/data/model/__init__.py pandas-ml-utils/pandas_ml_utils_test/ml/model/__init__.py

@@ -4,7 +4,7 @@
 from typing import Tuple
 
 from pandas_ml_common import pd, np, naive_splitter, random_splitter
-from pandas_ml_utils import FeaturesAndLabels, Model
+from pandas_ml_utils import FeaturesAndLabels, Model, SubModelFeature
 from pandas_ml_utils.ml.model.base_model import AutoEncoderModel
 
 
@@ -229,6 +229,38 @@ def test_no_test_data(self):
         self.assertEqual(len(fit.training_summary.df), len(df))
         self.assertEqual(len(fit.test_summary.df), 0)
 
+    def test_stacked_models(self):
+        """given some toy classification data"""
+        df = pd.DataFrame({
+            "a": [1, 0, 1, 0, 1, 0, 1, 0, ],
+            "b": [0, 0, 1, 1, 0, 0, 1, 1, ],
+            "c": [1, 0, 0, 1, 1, 0, 0, 1, ]
+        })
+
+        """and a model"""
+        model = self.provide_classification_model(
+            FeaturesAndLabels(
+                features=[
+                    "a",
+                    SubModelFeature("b", self.provide_classification_model(
+                        FeaturesAndLabels(features=["a", "b"], labels=["c"], label_type=int)
+                    ))
+                ],
+                labels=["c"],
+                label_type=int
+            )
+        )
+
+        with self.assertLogs(level='INFO') as cm:
+            with df.model() as m:
+                m.fit(model)
+
+            self.assertIn("INFO:pandas_ml_utils.ml.model.base_model:fitting submodel: b", cm.output[0])
+            self.assertIn(
+                "INFO:pandas_ml_utils.ml.model.base_model:fitted submodel",
+                [s for s in cm.output if s.startswith("INFO:pandas_ml_utils.ml.model.base_model:fitted")][0]
+            )
+
 
     # Abstract methods
     def provide_batch_size_and_epoch(self) -> Tuple[int, int]:

pandas-ml-utils/pandas_ml_utils_test/ml/data/model/test_abstract_model.py pandas-ml-utils/pandas_ml_utils_test/ml/model/test_abstract_model.py

@@ -5,7 +5,7 @@
 from sklearn.datasets import make_regression, make_classification
 from pandas_ml_common import np, pd, naive_splitter, stratified_random_splitter
 from pandas_ml_utils.ml.model.base_model import AutoEncoderModel
-from pandas_ml_utils_test.ml.data.model.test_abstract_model import TestAbstractModel
+from pandas_ml_utils_test.ml.model.test_abstract_model import TestAbstractModel
 from pandas_ml_utils import SkModel, SkAutoEncoderModel, FeaturesAndLabels, ClassificationSummary, RegressionSummary
 from pandas_ml_utils_test.config import DF_NOTES