Remove extra naive pipelines (#4142)

* Random Forest - Two Naive

* Elastic Net - Two Naive

* XGBoost - Two Naive

* 1 Naive RF - XGB Feat Select

* 1 Naive RF - EN Feat Select

* Revert feature selector changes

* Updated tests

* Set timeseries example to iterative algo

* Lint fix

* Lint notebook

docs/source/release_notes.rst

@@ -3,6 +3,7 @@ Release Notes
 **Future Releases**
     * Enhancements
         * Saved computed additional_objectives computed during search to AutoML object :pr:`4141`
+        * Remove extra naive pipelines :pr:`4142`
     * Fixes
         * Fixed usage of codecov after uploader deprecation :pr:`4144`
     * Changes

docs/source/user_guide/timeseries.ipynb

@@ -541,6 +541,7 @@
     "    problem_type=\"time series regression\",\n",
     "    max_batches=1,\n",
     "    problem_configuration=problem_config,\n",
+    "    automl_algorithm=\"iterative\",\n",
     "    allowed_model_families=[\n",
     "        \"xgboost\",\n",
     "        \"random_forest\",\n",

evalml/automl/automl_algorithm/default_algorithm.py

@@ -180,12 +180,10 @@ def num_pipelines_per_batch(self, batch_number):
     def _naive_estimators(self):
         if is_regression(self.problem_type):
             naive_estimators = [
-                "Elastic Net Regressor",
                 "Random Forest Regressor",
             ]
         else:
             naive_estimators = [
-                "Logistic Regression Classifier",
                 "Random Forest Classifier",
             ]
         estimators = [

evalml/tests/automl_tests/test_automl.py

@@ -261,9 +261,9 @@ def test_search_batch_times(caplog, X_y_binary, AutoMLTestEnv):
     assert isinstance(batch_times[3]["Total time of batch"], str)
 
     assert len(batch_times) == 3
-    assert len(batch_times[1]) == 3
-    assert len(batch_times[2]) == 3
-    assert len(batch_times[3]) == 7
+    assert len(batch_times[1]) == 2
+    assert len(batch_times[2]) == 2
+    assert len(batch_times[3]) == 8
 
     assert "Batch Time Stats" in out
     assert "Batch 1 time stats" in out
@@ -1781,14 +1781,14 @@ def test_pipelines_in_batch_return_nan(
 ):
     X, y = X_y_binary
     mock_rankings.side_effect = [
-        make_mock_rankings([0, 0, 0]),  # first batch
-        make_mock_rankings([0, 0, 0, 0, np.nan]),  # second batch
-        make_mock_rankings([0, 0, 0, 0, np.nan, np.nan, np.nan]),
+        make_mock_rankings([0, 0]),  # first batch
+        make_mock_rankings([0, 0, 0, np.nan]),  # second batch
+        make_mock_rankings([0, 0, 0, np.nan, np.nan, np.nan]),
     ]  # third batch, should raise error
     mock_full_rankings.side_effect = [
-        make_mock_rankings([0, 0, 0]),  # first batch
-        make_mock_rankings([0, 0, 0, 0, np.nan]),  # second batch
-        make_mock_rankings([0, 0, 0, 0, np.nan, np.nan, np.nan]),
+        make_mock_rankings([0, 0]),  # first batch
+        make_mock_rankings([0, 0, 0, np.nan]),  # second batch
+        make_mock_rankings([0, 0, 0, np.nan, np.nan, np.nan]),
     ]  # third batch, should raise error
     mock_next_batch.side_effect = [
         [
@@ -1834,14 +1834,14 @@ def test_pipelines_in_batch_return_none(
 ):
     X, y = X_y_binary
     mock_rankings.side_effect = [
-        make_mock_rankings([0, 0, 0]),  # first batch
-        make_mock_rankings([0, 0, 0, 0, None]),  # second batch
-        make_mock_rankings([0, 0, 0, 0, None, None, None]),
+        make_mock_rankings([0, 0]),  # first batch
+        make_mock_rankings([0, 0, 0, None]),  # second batch
+        make_mock_rankings([0, 0, 0, None, None, None]),
     ]  # third batch, should raise error
     mock_full_rankings.side_effect = [
-        make_mock_rankings([0, 0, 0]),  # first batch
-        make_mock_rankings([0, 0, 0, 0, None]),  # second batch
-        make_mock_rankings([0, 0, 0, 0, None, None, None]),
+        make_mock_rankings([0, 0]),  # first batch
+        make_mock_rankings([0, 0, 0, None]),  # second batch
+        make_mock_rankings([0, 0, 0, None, None, None]),
     ]  # third batch, should raise error
     mock_next_batch.side_effect = [
         [
@@ -2588,7 +2588,7 @@ def test_max_batches_plays_nice_with_other_stopping_criteria(AutoMLTestEnv, X_y_
     with env.test_context(score_return_value={"Log Loss Binary": 0.3}):
         automl.search()
 
-    assert len(automl.results["pipeline_results"]) == 3
+    assert len(automl.results["pipeline_results"]) == 2
 
 
 @pytest.mark.parametrize("max_batches", [-1, -10, -np.inf])

evalml/tests/automl_tests/test_default_algorithm.py

@@ -140,15 +140,15 @@ def test_default_algorithm(
     problem_type = automl_type
     sampler_name = None
     algo = DefaultAlgorithm(X, y, problem_type, sampler_name, ensembling=True)
-    naive_model_families = set([ModelFamily.LINEAR_MODEL, ModelFamily.RANDOM_FOREST])
+    naive_model_families = set([ModelFamily.RANDOM_FOREST])
 
     first_batch = algo.next_batch()
-    assert len(first_batch) == 2
+    assert len(first_batch) == 1
     assert {p.model_family for p in first_batch} == naive_model_families
     add_result(algo, first_batch)
 
     second_batch = algo.next_batch()
-    assert len(second_batch) == 2
+    assert len(second_batch) == 1
     assert {p.model_family for p in second_batch} == naive_model_families
     for pipeline in second_batch:
         assert pipeline.get_component(fs)
@@ -600,10 +600,10 @@ def test_default_algorithm_time_series(
         sampler_name,
         search_parameters=search_parameters,
     )
-    naive_model_families = set([ModelFamily.LINEAR_MODEL, ModelFamily.RANDOM_FOREST])
+    naive_model_families = set([ModelFamily.RANDOM_FOREST])
 
     first_batch = algo.next_batch()
-    assert len(first_batch) == 2 if problem_type != "time series regression" else 4
+    assert len(first_batch) == 1 if problem_type != "time series regression" else 4
     assert {p.model_family for p in first_batch} == naive_model_families
     for pipeline in first_batch:
         assert pipeline.parameters["pipeline"] == search_parameters["pipeline"]
@@ -680,10 +680,10 @@ def test_default_algorithm_time_series_known_in_advance(
         sampler_name,
         search_parameters=search_parameters,
     )
-    naive_model_families = set([ModelFamily.LINEAR_MODEL, ModelFamily.RANDOM_FOREST])
+    naive_model_families = set([ModelFamily.RANDOM_FOREST])
 
     first_batch = algo.next_batch()
-    assert len(first_batch) == 2
+    assert len(first_batch) == 1
     assert {p.model_family for p in first_batch} == naive_model_families
     for pipeline in first_batch:
         assert (