Delete baseline pipeline classes (#2202)

* init * fixed baseline for binary * convert over multiclass * fix tests * update regression pipelines * comment out time series * oops fix import * fix more imports * delete and cleanup * add test and fix input feature names
alteryx · May 2, 2021 · 2f7f653 · 2f7f653
1 parent 9d108d7
commit 2f7f653
Show file tree

Hide file tree

Showing 19 changed files with 148 additions and 636 deletions.
diff --git a/docs/source/api_reference.rst b/docs/source/api_reference.rst
@@ -151,30 +151,6 @@ Pipeline Base Classes
     TimeSeriesMulticlassClassificationPipeline
     TimeSeriesRegressionPipeline
 
-Classification Pipelines
-~~~~~~~~~~~~~~~~~~~~~~~~
-.. autosummary::
-    :toctree: generated
-    :template: pipeline_class.rst
-    :nosignatures:
-
-    BaselineBinaryPipeline
-    BaselineMulticlassPipeline
-    ModeBaselineBinaryPipeline
-    ModeBaselineMulticlassPipeline
-
-Regression Pipelines
-~~~~~~~~~~~~~~~~~~~~
-
-.. autosummary::
-    :toctree: generated
-    :template: pipeline_class.rst
-    :nosignatures:
-
-    BaselineRegressionPipeline
-    MeanBaselineRegressionPipeline
-    TimeSeriesBaselineRegressionPipeline
-
 
 .. currentmodule:: evalml.pipelines.utils
 

diff --git a/docs/source/release_notes.rst b/docs/source/release_notes.rst
@@ -10,6 +10,7 @@ Release Notes
     * Fixes
         * Fixed partial dependence not respecting grid resolution parameter for numerical features :pr:`2180`
     * Changes
+        * Deleted baseline pipeline classes :pr:`2202`
         * Reverting user specified date feature PR :pr:`2155` until `pmdarima` installation fix is found :pr:`2214`
         * Updated pipeline API to accept component graph and other class attributes as instance parameters. Old pipeline API still works but will not be supported long-term. :pr:`2091`
     * Documentation Changes
@@ -24,6 +25,7 @@ Release Notes
 .. warning::
 
     **Breaking Changes**
+        * All baseline pipeline classes (``BaselineBinaryPipeline``, ``BaselineMulticlassPipeline``, ``BaselineRegressionPipeline``, etc.) have been deleted :pr:`2202`
         * Updated pipeline API to accept component graph and other class attributes as instance parameters. Old pipeline API still works but will not be supported long-term. Pipelines can now be initialized by specifying the component graph as the first parameter, and then passing in optional arguments such as ``custom_name``, ``parameters``, etc. For example, ``BinaryClassificationPipeline(["Random Forest Classifier"], parameters={})``.  :pr:`2091`
 
 **v0.23.0 Apr. 20, 2021**

diff --git a/evalml/automl/automl_search.py b/evalml/automl/automl_search.py
@@ -34,13 +34,13 @@
     get_objective
 )
 from evalml.pipelines import (
-    MeanBaselineRegressionPipeline,
-    ModeBaselineBinaryPipeline,
-    ModeBaselineMulticlassPipeline,
+    BinaryClassificationPipeline,
+    MulticlassClassificationPipeline,
     PipelineBase,
-    TimeSeriesBaselineBinaryPipeline,
-    TimeSeriesBaselineMulticlassPipeline,
-    TimeSeriesBaselineRegressionPipeline
+    RegressionPipeline,
+    TimeSeriesBinaryClassificationPipeline,
+    TimeSeriesMulticlassClassificationPipeline,
+    TimeSeriesRegressionPipeline
 )
 from evalml.pipelines.components.utils import get_estimators
 from evalml.pipelines.utils import make_pipeline
@@ -648,25 +648,38 @@ def _validate_problem_type(self):
             if pipeline.problem_type != self.problem_type:
                 raise ValueError("Given pipeline {} is not compatible with problem_type {}.".format(pipeline.name, self.problem_type.value))
 
-    def _add_baseline_pipelines(self):
-        """Fits a baseline pipeline to the data.
-
-        This is the first pipeline fit during search.
-        """
+    def _get_baseline_pipeline(self):
+        """Creates a baseline pipeline instance."""
         if self.problem_type == ProblemTypes.BINARY:
-            baseline = ModeBaselineBinaryPipeline(parameters={})
+            baseline = BinaryClassificationPipeline(component_graph=["Baseline Classifier"],
+                                                    custom_name="Mode Baseline Binary Classification Pipeline",
+                                                    custom_hyperparameters={"strategy": ["mode"]})
         elif self.problem_type == ProblemTypes.MULTICLASS:
-            baseline = ModeBaselineMulticlassPipeline(parameters={})
+            baseline = MulticlassClassificationPipeline(component_graph=["Baseline Classifier"],
+                                                        custom_name="Mode Baseline Multiclass Classification Pipeline",
+                                                        custom_hyperparameters={"strategy": ["mode"]})
         elif self.problem_type == ProblemTypes.REGRESSION:
-            baseline = MeanBaselineRegressionPipeline(parameters={})
+            baseline = RegressionPipeline(component_graph=["Baseline Regressor"],
+                                          custom_name="Mean Baseline Regression Pipeline",
+                                          custom_hyperparameters={"strategy": ["mean"]})
         else:
-            pipeline_class = {ProblemTypes.TIME_SERIES_REGRESSION: TimeSeriesBaselineRegressionPipeline,
-                              ProblemTypes.TIME_SERIES_MULTICLASS: TimeSeriesBaselineMulticlassPipeline,
-                              ProblemTypes.TIME_SERIES_BINARY: TimeSeriesBaselineBinaryPipeline}[self.problem_type]
+            pipeline_class, pipeline_name = {ProblemTypes.TIME_SERIES_REGRESSION: (TimeSeriesRegressionPipeline, "Time Series Baseline Regression Pipeline"),
+                                             ProblemTypes.TIME_SERIES_MULTICLASS: (TimeSeriesMulticlassClassificationPipeline, "Time Series Baseline Multiclass Pipeline"),
+                                             ProblemTypes.TIME_SERIES_BINARY: (TimeSeriesBinaryClassificationPipeline, "Time Series Baseline Binary Pipeline")}[self.problem_type]
             gap = self.problem_configuration['gap']
             max_delay = self.problem_configuration['max_delay']
-            baseline = pipeline_class(parameters={"pipeline": {"gap": gap, "max_delay": max_delay},
+            baseline = pipeline_class(component_graph=["Time Series Baseline Estimator"],
+                                      custom_name=pipeline_name,
+                                      parameters={"pipeline": {"gap": gap, "max_delay": max_delay},
                                                   "Time Series Baseline Estimator": {"gap": gap, "max_delay": max_delay}})
+        return baseline
+
+    def _add_baseline_pipelines(self):
+        """Fits a baseline pipeline to the data.
+
+        This is the first pipeline fit during search.
+        """
+        baseline = self._get_baseline_pipeline()
         self._pre_evaluation_callback(baseline)
         logger.info(f"Evaluating Baseline Pipeline: {baseline.name}")
         computation = self._engine.submit_evaluation_job(self.automl_config, baseline, self.X_train, self.y_train)

diff --git a/evalml/pipelines/__init__.py b/evalml/pipelines/__init__.py
@@ -46,15 +46,4 @@
     TimeSeriesBinaryClassificationPipeline,
     TimeSeriesMulticlassClassificationPipeline
 )
-from .classification import (
-    BaselineBinaryPipeline,
-    BaselineMulticlassPipeline,
-    ModeBaselineBinaryPipeline,
-    ModeBaselineMulticlassPipeline
-)
 from .time_series_regression_pipeline import TimeSeriesRegressionPipeline
-from .regression import (
-    BaselineRegressionPipeline,
-    MeanBaselineRegressionPipeline,
-)
-from .time_series_baselines import TimeSeriesBaselineRegressionPipeline, TimeSeriesBaselineBinaryPipeline, TimeSeriesBaselineMulticlassPipeline
diff --git a/evalml/pipelines/classification/__init__.py b/evalml/pipelines/classification/__init__.py
diff --git a/evalml/pipelines/classification/baseline_binary.py b/evalml/pipelines/classification/baseline_binary.py
diff --git a/evalml/pipelines/classification/baseline_multiclass.py b/evalml/pipelines/classification/baseline_multiclass.py
diff --git a/evalml/pipelines/component_graph.py b/evalml/pipelines/component_graph.py
@@ -162,7 +162,9 @@ def _fit_transform_features_helper(self, needs_fitting, X, y=None):
             ww.DataTable: Transformed values.
         """
         if len(self.compute_order) <= 1:
-            return infer_feature_types(X)
+            X = infer_feature_types(X)
+            self.input_feature_names.update({self.compute_order[0]: list(X.columns)})
+            return X
         component_outputs = self._compute_features(self.compute_order[:-1], X, y=y, fit=needs_fitting)
         final_component_inputs = []
         for parent in self.get_parents(self.compute_order[-1]):

diff --git a/evalml/pipelines/regression/__init__.py b/evalml/pipelines/regression/__init__.py
diff --git a/evalml/pipelines/regression/baseline_regression.py b/evalml/pipelines/regression/baseline_regression.py
diff --git a/evalml/pipelines/time_series_baselines.py b/evalml/pipelines/time_series_baselines.py