def test_removeCourse_courseNotExists(self):
"""
Test tries to remove a course that does not exist,
then makes sure that the other courses are not affected.
"""
allCourses = self.myCourseCatalog.searchCoursesThroughPartialName(" ")
TestCase.assertEqual(
self, len(allCourses), 3,
"There are not the correct amount of courses in the database")
self.myCourseCatalog.removeCourseWithSections("COEN", 341)
allCourses = self.myCourseCatalog.searchCoursesThroughPartialName(" ")
TestCase.assertLess(self, len(allCourses), 4,
"One or many courses have been added")
TestCase.assertGreater(self, len(allCourses), 2,
"One or many courses have been removed")
def assertMostlyAgrees(test_case: TestCase, expected_data: pd.Series, actual_data: pd.Series):
"""Checks to make sure that the expected data and the actual data are mostly
the same, i.e. they have mostly the same genes and for the genes they have
in common their spearman correlation is >0.99.
We are only checking for approximate equality because output varies slightly
between runs and between bioconductor versions."""
# Make sure that the genes haven't changed too drastically between runs.
# If this fails, it's probably not the end of the world but probably
# something we should know about.
test_case.assertGreater(
len(set(expected_data.index) & set(actual_data.index)),
0.95 * min(len(set(expected_data.index)), len(set(actual_data.index))),
)
expected_df = squish_duplicates(pd.DataFrame({"expected_values": expected_data}))
actual_df = squish_duplicates(pd.DataFrame({"actual_values": actual_data}))
(rho, _) = scipy.stats.spearmanr(expected_df.join(actual_df, how="inner"))
test_case.assertGreater(rho, 0.99)
