gitea_admin/sientia-dataops-model-manager
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: log regression metrics as parameters in Training class

Browse Source 
- Added a method to persist computed regression metrics (MSE, MAE, R²) as MLflow parameters during model training, enhancing model evaluation and tracking.
- Updated the Training class to log the equation path if available, improving artifact management.
This commit is contained in:
vitor-aignosi
2026-04-17 10:52:32 -03:00
parent b245801e09
commit 31e95cbdf8
11 changed files with 649 additions and 1167 deletions
Download Patch File Download Diff File Expand all files Collapse all files

78
tests/sientia/test_reports.py
View File

@@ -31,9 +31,8 @@ def test_load_html_from_file_success(tmp_path):
def test_load_html_from_file_missing_file():
result = reports.load_html_from_file('non-existent.html')
assert result is None
with pytest.raises(FileNotFoundError):
reports.load_html_from_file('non-existent.html')
def test_load_html_from_file_os_error(monkeypatch):
@@ -42,9 +41,8 @@ def test_load_html_from_file_os_error(monkeypatch):
monkeypatch.setattr('builtins.open', fake_open)
result = reports.load_html_from_file('path.html')
assert result is None
with pytest.raises(OSError, match='boom'):
reports.load_html_from_file('path.html')
def test_inject_content_replaces_section():
@@ -71,7 +69,7 @@ def test_inject_content_missing_section():
def test_reports_init_sets_defaults(stub_color_options):
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
assert report.metrics == []
assert isinstance(report.options, list) and len(report.options) == 1
@@ -89,7 +87,7 @@ def test_add_data_quality_section_without_run(monkeypatch, stub_color_options):
monkeypatch.setattr(reports, 'ConflictTargetMetric', lambda: 'conflict')
monkeypatch.setattr(reports, 'DatasetCorrelationsMetric', lambda: 'correlations')
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
report.add_data_quality_section(columns=['col'], run=False)
assert report.metrics[-4:] == [
@@ -120,7 +118,9 @@ def test_add_data_quality_section_with_run(monkeypatch, tmp_path, stub_color_opt
ReportMock = MagicMock(return_value=report_instance)
monkeypatch.setattr(reports, 'Report', ReportMock)
report = reports.Reports(reference_data='ref', current_data='cur', base_path=str(tmp_path))
report = reports.Reports(
reference_data='ref', current_data='cur', target_name='target', base_path=str(tmp_path)
)
report.add_data_quality_section(columns=['c1'], run=True)
assert report.metrics[-4:] == [summary, column_metrics, conflict, correlations]
@@ -153,7 +153,7 @@ def test_add_data_quality_section_run_without_base_path(monkeypatch, stub_color_
ReportMock = MagicMock(return_value=report_instance)
monkeypatch.setattr(reports, 'Report', ReportMock)
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
report.add_data_quality_section(run=True)
assert report.sections['data_quality'] == {'result': 'quality'}
@@ -170,7 +170,9 @@ def test_add_data_drift_section_paths(monkeypatch, tmp_path, stub_color_options)
ReportMock = MagicMock(return_value=report_instance)
monkeypatch.setattr(reports, 'Report', ReportMock)
report = reports.Reports(reference_data='ref', current_data='cur', base_path=str(tmp_path))
report = reports.Reports(
reference_data='ref', current_data='cur', target_name='target', base_path=str(tmp_path)
)
report.add_data_drift_section(columns=['c1'], run=False)
assert report.metrics[-1] == drift_instances[0]
assert 'data_drift' not in report.sections
@@ -192,7 +194,7 @@ def test_add_data_drift_section_run_without_base_path(monkeypatch, stub_color_op
ReportMock = MagicMock(return_value=report_instance)
monkeypatch.setattr(reports, 'Report', ReportMock)
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
report.add_data_drift_section(run=True)
assert report.sections['data_drift'] == {'result': 'drift'}
@@ -216,7 +218,9 @@ def test_add_regression_section(monkeypatch, tmp_path, stub_color_options):
ReportMock = MagicMock(return_value=report_instance)
monkeypatch.setattr(reports, 'Report', ReportMock)
report = reports.Reports(reference_data='ref', current_data='cur', base_path=str(tmp_path))
report = reports.Reports(
reference_data='ref', current_data='cur', target_name='target', base_path=str(tmp_path)
)
report.add_regression_section(run=False)
assert report.metrics[-7:] == regression_metrics
@@ -225,7 +229,11 @@ def test_add_regression_section(monkeypatch, tmp_path, stub_color_options):
report.add_regression_section(run=True)
assert report.sections['regression'] == {'result': 'regression'}
ReportMock.assert_called_with(metrics=regression_metrics, options=report.options)
report_instance.run.assert_called_with(reference_data='ref', current_data='cur')
report_instance.run.assert_called_with(
reference_data='ref',
current_data='cur',
column_mapping=report_instance.run.call_args.kwargs['column_mapping'],
)
report_instance.save_html.assert_called_with(os.path.join(str(tmp_path), 'regression.html'))
@@ -246,7 +254,7 @@ def test_add_regression_section_run_without_base_path(monkeypatch, stub_color_op
ReportMock = MagicMock(return_value=report_instance)
monkeypatch.setattr(reports, 'Report', ReportMock)
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
report.add_regression_section(run=True)
assert report.sections['regression'] == {'result': 'reg'}
@@ -262,7 +270,7 @@ def test_set_color_options_appends(monkeypatch):
monkeypatch.setattr(reports, 'ColorOptions', color_options_mock)
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
report.set_color_options(primary_color='#111', secondary_color='#222')
assert len(report.options) == 2
@@ -272,12 +280,24 @@ def test_set_color_options_appends(monkeypatch):
def test_save_all_sections_html_requires_base_path(stub_color_options):
report = reports.Reports(reference_data='ref', current_data='cur')
report = reports.Reports(reference_data='ref', current_data='cur', target_name='target')
with pytest.raises(ValueError):
report.save_all_sections_html('output/report.html')
def test_save_all_sections_html_requires_template_path(stub_color_options, tmp_path):
report = reports.Reports(
reference_data='ref',
current_data='cur',
target_name='target',
base_path=str(tmp_path),
)
with pytest.raises(ValueError, match='template_path is required'):
report.save_all_sections_html('output/report.html')
def test_save_all_sections_html_writes_output(tmp_path, stub_color_options):
base_dir = tmp_path / 'templates'
base_dir.mkdir()
@@ -289,7 +309,13 @@ def test_save_all_sections_html_writes_output(tmp_path, stub_color_options):
(base_dir / 'data_quality.html').write_text('<p>Quality</p>', encoding='utf-8')
(base_dir / 'regression.html').write_text('<p>Regression</p>', encoding='utf-8')
report = reports.Reports(reference_data='ref', current_data='cur', base_path=str(base_dir))
report = reports.Reports(
reference_data='ref',
current_data='cur',
target_name='target',
base_path=str(base_dir),
template_path=str(base_dir),
)
output_path = tmp_path / 'reports' / 'combined.html'
report.save_all_sections_html(str(output_path))
@@ -313,7 +339,13 @@ def test_save_all_sections_html_creates_directory(monkeypatch, tmp_path, stub_co
(base_dir / 'regression.html').write_text('<p>Regression</p>', encoding='utf-8')
make_dirs_called = []
report = reports.Reports(reference_data='ref', current_data='cur', base_path=str(base_dir))
report = reports.Reports(
reference_data='ref',
current_data='cur',
target_name='target',
base_path=str(base_dir),
template_path=str(base_dir),
)
output_path = tmp_path / 'nested' / 'report.html'
output_dir = str(output_path.parent)
@@ -356,7 +388,13 @@ def test_save_all_sections_html_no_directory_needed(monkeypatch, tmp_path, stub_
monkeypatch.setattr(os, 'makedirs', fake_makedirs)
monkeypatch.chdir(tmp_path)
report = reports.Reports(reference_data='ref', current_data='cur', base_path=str(base_dir))
report = reports.Reports(
reference_data='ref',
current_data='cur',
target_name='target',
base_path=str(base_dir),
template_path=str(base_dir),
)
report.save_all_sections_html('report.html')
assert mk_calls == []