adding pytest mark for db requirement

fmohr · Oct 8, 2024 · 71504fa · 71504fa
1 parent 90120bf
commit 71504fa
Show file tree

Hide file tree

Showing 2 changed files with 34 additions and 25 deletions.
diff --git a/publications/2023-neurips/pytest.ini b/publications/2023-neurips/pytest.ini
@@ -1,4 +1,6 @@
 [pytest]
-norecursedirs = .git 
+norecursedirs = .git
+markers =
+    db: marks to define a test that requires a local database from LCDB data.
 filterwarnings =
     ignore:The objective has been evaluated at this point before.:UserWarning
diff --git a/publications/2023-neurips/test/test_exctraction.py b/publications/2023-neurips/test/test_exctraction.py
@@ -1,23 +1,28 @@
-from lcdb.db import LCDB
-
-from parameterized import parameterized
+import pytest
 import unittest
 
 import numpy as np
 from lcdb.analysis.util import LearningCurveExtractor, merge_curves
+from lcdb.db import LCDB
+from parameterized import parameterized
 
 
+@pytest.mark.db
 class TestExtractors(unittest.TestCase):
 
-    @parameterized.expand([
-        (6, "lcdb.workflow.sklearn.KNNWorkflow", 0, 0, 42),
-        (3, "lcdb.workflow.sklearn.LibLinearWorkflow", 0, 0, 42),
-        (3, "lcdb.workflow.sklearn.LibSVMWorkflow", 0, 0, 42),
-        (6, "lcdb.workflow.sklearn.TreesEnsembleWorkflow", 0, 1, 42)
-    ])
-    def test_learning_curve_extraction(self, openmlid, workflow, val_seed, test_seed, workflow_seed):
+    @parameterized.expand(
+        [
+            (6, "lcdb.workflow.sklearn.KNNWorkflow", 0, 0, 42),
+            (3, "lcdb.workflow.sklearn.LibLinearWorkflow", 0, 0, 42),
+            (3, "lcdb.workflow.sklearn.LibSVMWorkflow", 0, 0, 42),
+            (6, "lcdb.workflow.sklearn.TreesEnsembleWorkflow", 0, 1, 42),
+        ]
+    )
+    def test_learning_curve_extraction(
+        self, openmlid, workflow, val_seed, test_seed, workflow_seed
+    ):
 
-        metrics = ["error_rate"]#, "balanced_error_rate"]
+        metrics = ["error_rate"]  # , "balanced_error_rate"]
 
         lcdb = LCDB()
         df = lcdb.query(
@@ -29,11 +34,10 @@ def test_learning_curve_extraction(self, openmlid, workflow, val_seed, test_seed
             workflow_seeds=[workflow_seed],
             processors={
                 "learning_curve": LearningCurveExtractor(
-                    metrics=metrics,
-                    folds=["train", "val", "test", "oob"]
+                    metrics=metrics, folds=["train", "val", "test", "oob"]
                 )
             },
-            show_progress=True
+            show_progress=True,
         )
 
         oob_fold_expected = "TreesEnsembleWorkflow" in workflow
@@ -53,12 +57,14 @@ def test_learning_curve_extraction(self, openmlid, workflow, val_seed, test_seed
 
         self.assertTrue(not oob_fold_expected or num_oob_not_nan > 0)
 
-    @parameterized.expand([
-        #(6, "lcdb.workflow.sklearn.KNNWorkflow"),
-        (3, "lcdb.workflow.sklearn.LibLinearWorkflow"),
-        (3, "lcdb.workflow.sklearn.LibSVMWorkflow"),
-        #(6, "lcdb.workflow.sklearn.TreesEnsembleWorkflow")
-    ])
+    @parameterized.expand(
+        [
+            # (6, "lcdb.workflow.sklearn.KNNWorkflow"),
+            (3, "lcdb.workflow.sklearn.LibLinearWorkflow"),
+            (3, "lcdb.workflow.sklearn.LibSVMWorkflow"),
+            # (6, "lcdb.workflow.sklearn.TreesEnsembleWorkflow")
+        ]
+    )
     def test_learning_curve_grouping_after_extraction(self, openmlid, workflow):
 
         lcdb = LCDB()
@@ -71,16 +77,17 @@ def test_learning_curve_grouping_after_extraction(self, openmlid, workflow):
             validation_seeds=validation_seeds,
             processors={
                 "learning_curve": LearningCurveExtractor(
-                    metrics=["error_rate"],
-                    folds=["train", "val", "test", "oob"]
+                    metrics=["error_rate"], folds=["train", "val", "test", "oob"]
                 )
             },
-            show_progress=True
+            show_progress=True,
         )
 
         if df is not None:
 
             config_cols = [c for c in df.columns if c.startswith("p:")]
             len_before = len(df)
-            len_after = len(df.groupby(config_cols).agg({"learning_curve": merge_curves}))
+            len_after = len(
+                df.groupby(config_cols).agg({"learning_curve": merge_curves})
+            )
             self.assertEqual(len_before, len_after * len(validation_seeds))