batch error logs

setup.py

@@ -50,6 +50,7 @@
         "tqdm",
         "layered_config_tree>=2.1.0",
         "loguru",
+        "pytest_check",
         # type stubs
         "pandas-stubs",
         "types-PyYAML",

tests/integration/release/test_release.py

@@ -5,6 +5,7 @@
 import pandas as pd
 import pytest
 from _pytest.fixtures import FixtureRequest
+from pytest_check import check  # type: ignore
 from vivarium_testing_utils import FuzzyChecker
 
 from pseudopeople.dataset import Dataset
@@ -60,11 +61,12 @@ def test_unnoised_id_cols(dataset_name: str, request: FixtureRequest) -> None:
     original = initialize_dataset_with_sample(dataset_name)
     noised_data = request.getfixturevalue("noised_data")
     check_noised, check_original, _ = _get_common_datasets(original, noised_data)
-    assert (
-        (
-            check_original.reset_index()[unnoised_id_cols]
-            == check_noised.reset_index()[unnoised_id_cols]
+    with check:
+        assert (
+            (
+                check_original.reset_index()[unnoised_id_cols]
+                == check_noised.reset_index()[unnoised_id_cols]
+            )
+            .all()
+            .all()
         )
-        .all()
-        .all()
-    )

tests/integration/release/test_runner.py

@@ -19,7 +19,7 @@
 )
 def test_release_tests(pytest_args: list[str]) -> None:
     os.chdir(Path(__file__).parent)  # need this to access cli options from conftest.py
-    base_cmd = ["pytest", "--release", "test_release.py"]
+    base_cmd = ["pytest", "--release", "test_release.py", "--check-max-tb=1000"]
     cmd = base_cmd + pytest_args
     result = subprocess.run(cmd, capture_output=True, text=True)
     assert result.returncode == 0

tests/utilities.py

@@ -8,6 +8,7 @@
 import numpy as np
 import numpy.typing as npt
 import pandas as pd
+from pytest_check import check  # type: ignore
 from vivarium_testing_utils import FuzzyChecker
 
 from pseudopeople.configuration import Keys, get_configuration
@@ -34,7 +35,8 @@ def run_column_noising_tests(
 
         # Check that originally missing data remained missing
         originally_missing_idx = check_original.index[check_original[col.name].isna()]
-        assert check_noised.loc[originally_missing_idx, col.name].isna().all()
+        with check:
+            assert check_noised.loc[originally_missing_idx, col.name].isna().all()
 
         # Check for noising where applicable
         to_compare_idx = shared_idx.difference(originally_missing_idx)
@@ -43,7 +45,8 @@ def run_column_noising_tests(
                 check_original.loc[to_compare_idx, col.name].values
                 != check_noised.loc[to_compare_idx, col.name].values
             )
-            assert different_check.any()
+            with check:
+                assert different_check.any()
 
             noise_level = different_check.sum()
 
@@ -64,7 +67,8 @@ def run_column_noising_tests(
                 == check_noised.loc[to_compare_idx, col.name].values
             )
 
-            assert same_check.all()
+            with check:
+                assert same_check.all()
 
 
 def run_omit_row_or_do_not_respond_tests(
@@ -87,15 +91,19 @@ def run_omit_row_or_do_not_respond_tests(
     ]:
         # Census and household surveys have do_not_respond and omit_row.
         # For all other datasets they are mutually exclusive
-        assert len(noise_types) == 2
+        with check:
+            assert len(noise_types) == 2
     else:
-        assert len(noise_types) < 2
+        with check:
+            assert len(noise_types) < 2
     if not noise_types:  # Check that there are no missing indexes
-        assert noised_data.index.symmetric_difference(original_data.index).empty
+        with check:
+            assert noised_data.index.symmetric_difference(original_data.index).empty
     else:  # Check that there are some omissions
         # TODO: assert levels are as expected
-        assert noised_data.index.difference(original_data.index).empty
-        assert not original_data.index.difference(noised_data.index).empty
+        with check:
+            assert noised_data.index.difference(original_data.index).empty
+            assert not original_data.index.difference(noised_data.index).empty
 
 
 def validate_column_noise_level(
@@ -158,7 +166,8 @@ def validate_column_noise_level(
                     [1 - p for p in token_probability]
                 )
             else:
-                assert isinstance(tokens_per_string, pd.Series)
+                with check:
+                    assert isinstance(tokens_per_string, pd.Series)
                 avg_probability_any_token_noised = (
                     1 - (1 - token_probability) ** tokens_per_string
                 ).mean()