Build: #6 failed
Job: Test Many Linux 2.28 failed
uid a002 x85c183 x36f SPw15 23 PPr regression: Test case result
The below summarizes the result of the test " uid a002 x85c183 x36f SPw15 23 PPr regression" in build 6 of Pipeline - Pipeline Main with Casa 6.6.6 test - cvpost - release-6.6.6 - Test Many Linux 2.28.
- Description
- uid a002 x85c183 x36f SPw15 23 PPr regression
- Test class
- pipeline.infrastructure.utils.regression-tester
- Method
- test_uid___A002_X85c183_X36f_SPW15_23__PPR__regression
- Duration
- 16 mins
- Status
- Failed (Existing Failure)
Error Log
Failed: Failed to match 2 result values within tolerances :
s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.metric.SingleDishImageContamination
values differ by > a relative difference of 1e-07
expected: True
new: False
diff: 1
percent_diff: 100.0%
s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.score.SingleDishImageContamination
values differ by > a relative difference of 1e-07
expected: 0.65
new: 1.0
diff: -0.35
percent_diff: -53.84615384615385%
Worst absolute diff, s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.metric.SingleDishImageContamination: 1
Worst percentage diff, s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.metric.SingleDishImageContamination: 100.0%
@pytest.mark.fast
@pytest.mark.alma
def test_uid___A002_X85c183_X36f_SPW15_23__PPR__regression():
"""Run ALMA single-dish restoredata regression on the observation data of M100.
Dataset: uid___A002_X85c183_X36f_SPW15_23
Expected results version: casa-6.2.1-2-pipeline-2021.2.0.94
"""
input_dir = 'pl-regressiontest/uid___A002_X85c183_X36f_SPW15_23'
pr = PipelineRegression(input_dir=input_dir,
visname=['uid___A002_X85c183_X36f_SPW15_23.ms'],
expectedoutput_dir=('pl-regressiontest/uid___A002_X85c183_X36f_SPW15_23'))
# copy files use restore task into products folder
input_products = casa_tools.utils.resolve(f'{input_dir}/products')
shutil.copytree(input_products, f'{pr.output_dir}/products')
> pr.run(ppr=f'{input_dir}/PPR.xml')
pipeline/infrastructure/utils/regression-tester.py:526:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
pipeline/infrastructure/utils/regression-tester.py:224: in run
self.__compare_results(new_file, default_relative_tolerance)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = <pipeline.infrastructure.utils.regression-tester.PipelineRegression object at 0x7faada9a90a0>
new_file = 'uid___A002_X85c183_X36f_SPW15_23.ms.NEW.results.txt'
relative_tolerance = 1e-07
def __compare_results(self, new_file: str, relative_tolerance: float):
"""
Compare results between new one loaded from file and old one.
Args:
new_file : file path of new results
relative_tolerance : relative tolerance of output value
"""
with open(self.expectedoutput_file) as expected_fd, open(new_file) as new_fd:
expected_results = expected_fd.readlines()
new_results = new_fd.readlines()
errors = []
worst_diff = (0, 0)
worst_percent_diff = (0, 0)
for old, new in zip(expected_results, new_results):
try:
oldkey, oldval, tol = self.__sanitize_regression_string(old)
newkey, newval, _ = self.__sanitize_regression_string(new)
except ValueError as e:
errorstr = "The results: {0} could not be parsed. Error: {1}".format(new, str(e))
errors.append(errorstr)
continue
assert oldkey == newkey
tolerance = tol if tol else relative_tolerance
if newval is not None:
LOG.info(f'Comparing {oldval} to {newval} with a rel. tolerance of {tolerance}')
if oldval != pytest.approx(newval, rel=tolerance):
diff = oldval-newval
percent_diff = (oldval-newval)/oldval * 100
if abs(diff) > abs(worst_diff[0]):
worst_diff = diff, oldkey
if abs(percent_diff) > abs(worst_percent_diff[0]):
worst_percent_diff = percent_diff, oldkey
errorstr = f"{oldkey}\n\tvalues differ by > a relative difference of {tolerance}\n\texpected: {oldval}\n\tnew: {newval}\n\tdiff: {diff}\n\tpercent_diff: {percent_diff}%"
errors.append(errorstr)
elif oldval is not None:
# If only the new value is None, fail
errorstr = f"{oldkey}\n\tvalue is None\n\texpected: {oldval}\n\tnew: {newval}"
errors.append(errorstr)
else:
# If old and new values are both None, this is expected, so pass
LOG.info(f'Comparing {oldval} and {newval}... both values are None.')
[LOG.warning(x) for x in errors]
n_errors = len(errors)
if n_errors > 0:
summary_str = f"Worst absolute diff, {worst_diff[1]}: {worst_diff[0]}\nWorst percentage diff, {worst_percent_diff[1]}: {worst_percent_diff[0]}%"
errors.append(summary_str)
> pytest.fail("Failed to match {0} result value{1} within tolerance{1} :\n{2}".format(
n_errors, '' if n_errors == 1 else 's', '\n'.join(errors)), pytrace=True)
E Failed: Failed to match 2 result values within tolerances :
E s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.metric.SingleDishImageContamination
E values differ by > a relative difference of 1e-07
E expected: True
E new: False
E diff: 1
E percent_diff: 100.0%
E s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.score.SingleDishImageContamination
E values differ by > a relative difference of 1e-07
E expected: 0.65
E new: 1.0
E diff: -0.35
E percent_diff: -53.84615384615385%
E Worst absolute diff, s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.metric.SingleDishImageContamination: 1
E Worst percentage diff, s7.hsd_imaging.uid___A002_X85c183_X36f_SPW15_23.ms.atmcor.qa.metric.SingleDishImageContamination: 100.0%
pipeline/infrastructure/utils/regression-tester.py:290: Failed