def testYVariableError(self):
"""
Test exception when bad "y" variable name is provided.
"""
with self.assertRaises(combine_csv.CombineCSV.CombineError) as cerr:
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"bad_y54_name")
self.assertEqual(cerr.exception._name, "YVariableError")
def testBasenameError(self):
"""
Test exception when bad basename is provided.
"""
with self.assertRaises(combine_csv.CombineCSV.CombineError) as cerr:
df_test = combine_csv.CombineCSV('bad_basename_54',
"remove_me_54.csv",
"large_number")
self.assertEqual(cerr.exception._name, "BasenameError")
def testXVariableError(self):
"""
Test exception when bad "x" variable name is provided.
"""
with self.assertRaises(combine_csv.CombineCSV.CombineError) as cerr:
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"large_number",
x_varname='bad_x54_name')
self.assertEqual(cerr.exception._name, "XVariableError")
def testInconsistentError(self):
"""
Test exception when data rows are not consistent.
"""
with self.assertRaises(combine_csv.CombineCSV.CombineError) as cerr:
df_test = combine_csv.CombineCSV("{}54_bad_".format(
self.__basename),
"remove_me_54.csv",
"large_number",
x_varname='x')
self.assertEqual(cerr.exception._name, "InconsistentError")
def testStepBoundsError(self):
"""
Test exception when mismatch of steps are provided.
"""
with self.assertRaises(combine_csv.CombineCSV.CombineError) as cerr:
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"large_number",
lastn=2,
endt=1)
self.assertEqual(cerr.exception._name, "StepBoundsError")
def testBasic(self):
"""
Test basic usage with minimal options and headers written.
"""
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"large_number",
write_header=True)
self.assertTrue(df_test._ended)
gold_df = pandas.read_csv("{}/combine_basic.csv".format(
self.__goldpath))
self.assertTrue(
df_test._final_df.equals(gold_df),
msg="Pandas dataframe is different from gold CSV for basic usage.")
def testBasicX(self):
"""
Test basic usage with headers and a "x" variable name.
"""
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"large_number",
write_header=True,
x_varname='y')
self.assertTrue(df_test._ended)
gold_df = pandas.read_csv("{}/combine_basic_x.csv".format(
self.__goldpath))
self.assertTrue(
df_test._final_df.equals(gold_df),
msg=
"Pandas dataframe is different from gold CSV for x variable usage."
)
def testBilinear(self):
"""
Test bilinear usage.
"""
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"large_number",
write_header=True,
x_varname='y',
timefile=True,
bilinear=True)
self.assertTrue(df_test._ended)
gold_df = pandas.read_csv("{}/combine_bilinear.csv".format(
self.__goldpath))
self.assertTrue(
df_test._final_df.equals(gold_df),
msg=
"Pandas dataframe is different from gold CSV for bilinear usage.")
def testBilinear(self):
"""
Test bilinear usage.
"""
df_test = combine_csv.CombineCSV(self.__basename,
"remove_me_54.csv",
"large_number",
write_header=True,
x_varname='y',
timefile=True,
bilinear=True)
self.assertTrue(df_test._ended)
gold_df = pandas.read_csv("{}/combine_bilinear.csv".format(
self.__goldpath))
self.assertTrue(
df_test._final_df.shape == gold_df.shape,
msg=
"Pandas dataframe size is different from gold CSV for bilinear usage."
)
# We used to use DataFrame.equals for a comparison here, but because a
# newer version of Pandas resulted in a very small diff in the column
# data, we can't use it anymore. Therefore - we need to do a more fuzzy
# check on both the columns and th data.
data_difference = df_test._final_df.to_numpy() - gold_df.to_numpy()
column_difference = df_test._final_df.columns.to_numpy(dtype=numpy.double) \
- gold_df.columns.to_numpy(dtype=numpy.double)
self.assertTrue(
abs(data_difference.max()) < 1e-10,
msg="Pandas data is different from gold CSV for bilinear usage.")
self.assertTrue(
abs(column_difference.max()) < 1e-10,
msg=
"Pandas column data is different from gold CSV for bilinear usage."
)