def test__no_filt_or_useless_filt__no_change(self, runfilter_test_df):
# BEFORE: Copy dataframe
df_copy1 = runfilter_test_df.copy()
# Filter the copy of the dataframe with no filter
result_indices, filt_map = df_util.run_filters(df_copy1, {})
# AFTER:
# Confirm filt_map correct
assert len(filt_map) == 0
# Confirm dataframe is unchanged
assert df_copy1.equals(runfilter_test_df)
######
# BEFORE: Copy dataframes
df_copy2 = runfilter_test_df.copy()
# Filter the copy of the dataframe with useless filter
func_name = "useless"
result_indices, filt_map = df_util.run_filters(
df_copy1, {func_name: lambda df: (df == -1).any(axis='columns')})
# AFTER:
# Confirm filt_map correct
assert len(filt_map) == 1
assert func_name in filt_map
# Confirm dataframe is unchanged
assert df_copy2.equals(runfilter_test_df)
assert not any(filt_map[func_name])
assert not any(result_indices)
def test__two_filt_no_overlap__return_expected_results(
self, runfilter_test_df):
# Identify value whose row to filter out and the expected filtering indices
target_int1 = 0
target_int2 = runfilter_test_df.size - 1
expected_indices1 = (runfilter_test_df == target_int1).any(
axis='columns')
expected_indices2 = (runfilter_test_df == target_int2).any(
axis='columns')
# BEFORE: Confirm dataframe contains the target int once
assert expected_indices1.sum() == 1
assert expected_indices2.sum() == 1
# Filter the dataframe
func_name1 = "Hello"
func_name2 = "World"
result_indices, filt_map = df_util.run_filters(
runfilter_test_df, {
func_name1: lambda df: (df == target_int1).any(axis='columns'),
func_name2: lambda df: (df == target_int2).any(axis='columns')
})
# AFTER:
# Confirm filt_map correct
assert len(filt_map) == 2
assert func_name1 in filt_map
assert func_name2 in filt_map
# Confirm result indices correct
assert all(result_indices == filt_map[func_name1]
| filt_map[func_name2])
assert result_indices.sum() == 2
def test__one_filt_duplicated__same_as_just_once(self, runfilter_test_df):
# Identify value whose row to filter out and the expected filtering indices
target_int = runfilter_test_df.size // 2
expected_indices = (runfilter_test_df == target_int).any(
axis='columns')
# BEFORE: Confirm dataframe contains the target int once
assert expected_indices.sum() == 1
# Filter the dataframe
func_name1 = "Hello, world"
func_name2 = "Goodbye, everyone"
result_indices, filt_map = df_util.run_filters(
runfilter_test_df, {
func_name1: lambda df: (df == target_int).any(axis='columns'),
func_name2: lambda df: (df == target_int).any(axis='columns')
})
# AFTER:
# Confirm filt_map correct
assert len(filt_map) == 2
assert func_name1 in filt_map
assert func_name2 in filt_map
# Confirm result indices correct
assert all(result_indices == filt_map[func_name1])
assert all(result_indices == filt_map[func_name2])
assert all(result_indices == expected_indices)
def test__failed_req_filt__throws_error(self, runfilter_test_df):
# Identify value whose row to filter out and the expected filtering indices
target_int = runfilter_test_df.size // 2
expected_indices = (runfilter_test_df == target_int).any(
axis='columns')
# BEFORE: Confirm dataframe contains the target int once and has the correct number of rows
assert expected_indices.sum() == 1
# Filter the dataframe
func_name = "Hello, world"
with pytest.raises(ZeroDivisionError):
result_indices, filt_map = df_util.run_filters(
runfilter_test_df, {func_name: lambda df: 1 / 0})
def test__empty_dataframe__no_change(self):
# Fill DF with numbers 0..N-1 where there are N cells
df = pd.DataFrame()
# BEFORE: Confirm dataframe copy contains the correct number of rows
assert len(df) == 0
# Filter the dataframe with useless filter
func_name = "useless"
result_indices, filt_map = df_util.run_filters(
df, {func_name: lambda df: (df == -1).any(axis='columns')})
# AFTER:
# Confirm filt_map correct
assert len(filt_map) == 1
assert func_name in filt_map
# Confirm dataframe is unchanged
assert len(df) == 0
assert not any(filt_map[func_name])
assert not any(result_indices)
def test__single_filt__return_expected_results(self, runfilter_test_df):
# Identify value whose row to filter out and the expected filtering indices
target_int = runfilter_test_df.size // 2
expected_indices = (runfilter_test_df == target_int).any(
axis='columns')
# BEFORE: Confirm dataframe contains the target int once and has the correct number of rows
assert expected_indices.sum() == 1
# Filter the dataframe
func_name = "Hello, world"
result_indices, filt_map = df_util.run_filters(
runfilter_test_df,
{func_name: lambda df: (df == target_int).any(axis='columns')})
# AFTER:
# Confirm filt_map correct
assert len(filt_map) == 1
assert func_name in filt_map
# Confirm result indices correct
assert all(result_indices == filt_map[func_name])
assert all(result_indices == expected_indices)