def test_update(self):
dict1 = ColumnDict()
stern_df = pd.read_csv(self.stern_map_path, delimiter='\t', header=0)
dict1.update(stern_df)
self.assertEqual(
len(dict1.value_info.keys()), 0,
"ColumnDict value_info should be empty if no value columns")
self.assertEqual(
len(dict1.categorical_info.keys()), len(stern_df.columns),
"ColumnDict categorical_info should have all the columns if no restrictions"
)
dict2 = ColumnDict(value_cols=['letter'], skip_cols=['event_type'])
dict2.update(stern_df)
self.assertEqual(
len(dict2.value_info.keys()), 1,
"ColumnDict value_info should have letter value column")
self.assertEqual(dict2.value_info['letter'], len(stern_df),
"ColumnDict value counts should length of column")
self.assertEqual(len(dict2.skip_cols), 1,
"ColumnDict should have one skip column")
self.assertEqual(
len(dict2.categorical_info.keys()),
len(stern_df.columns) - 2,
"ColumnDict categorical_info should have all the columns except value and skip columns"
)
dict2.update(stern_df)
self.assertEqual(
len(dict2.value_info.keys()), 1,
"ColumnDict value_info should have letter value column")
self.assertEqual(
dict2.value_info['letter'], 2 * len(stern_df),
"ColumnDict value counts should update by column length each time update is called"
)
def test_constructor(self):
dict1 = ColumnDict()
self.assertIsInstance(
dict1, ColumnDict,
"ColumnDict constructor is allowed to have no arguments")
self.assertFalse(dict1.value_info)
dict2 = ColumnDict(value_cols=['a', 'b', 'c'], name='baloney')
self.assertIsInstance(
dict2, ColumnDict,
"ColumnDict: multiple values are okay in constructor")
self.assertEqual(len(dict2.value_info.keys()), 3,
"ColumnDict should have keys for each value column")
def test_get_flattened(self):
t_map = ColumnDict()
t_map.update(self.stern_map_path)
df1 = t_map.get_flattened()
self.assertEqual(
len(df1), 137,
"Flattening should produce same number of items whether or not value used"
)
self.assertEqual(len(df1.columns), 2,
"A flattened map should have 2 columns")
t_map2 = ColumnDict(value_cols=['letter'], skip_cols=['type'])
t_map2.update(self.stern_map_path)
df2 = t_map2.get_flattened()
self.assertGreater(
len(df1), len(df2),
"Flattening of few columns produces smaller flattened array")
self.assertEqual(
len(df2.columns), 2,
"Flattening should produce 2 columns when value and skip columns")
def test_update_dict(self):
dict1 = ColumnDict()
dict2 = ColumnDict()
stern_df_map = pd.read_csv(self.stern_map_path,
delimiter='\t',
header=0)
dict1.update(stern_df_map)
dict2.update_dict(dict1)
dict2.update_dict(dict1)
dict1.update(stern_df_map)
cat_keys = dict1.categorical_info.keys()
for col in cat_keys:
self.assertTrue(col in dict2.categorical_info.keys())
key_dict1 = dict1.categorical_info[col]
key_dict2 = dict2.categorical_info[col]
for key, value in key_dict1.items():
self.assertEqual(
value, key_dict2[key],
"update_dict should give same values as update")
def test_print(self):
from io import StringIO
t_map = ColumnDict()
t_map.update(self.stern_map_path)
df = get_new_dataframe(self.stern_map_path)
t_map.update(self.stern_map_path)
self.assertEqual(
len(t_map.categorical_info.keys()), len(df.columns),
"ColumnDict should have all columns as categorical if no value or skip are given"
)
with mock.patch('sys.stdout', new=StringIO()):
t_map.print()
print("This should be eaten by the StringIO")
def test_update_dict_bad_skip_col(self):
stern_test3 = pd.read_csv(self.stern_test3_path,
delimiter='\t',
header=0)
dict1 = ColumnDict(skip_cols=['latency'])
dict1.update(stern_test3)
dict2 = ColumnDict(value_cols=['latency'])
dict2.update(stern_test3)
try:
dict2.update_dict(dict1)
except HedFileError:
pass
except Exception as ex:
self.fail(
f'update_dict threw the wrong exception {ex} when value columns should be categorical'
)
else:
self.fail(
'update_dict should have thrown a HedFileError when value columns should be categorical'
)
def make_combined_dicts(file_dict, skip_cols=None):
""" Return a combined dictionary of column information as we
Args:
file_dict (dict): Dictionary of file name keys and full path
skip_cols (list): Name of the column
Returns:
dict: A combined dictionary
"""
dicts_all = ColumnDict(skip_cols=skip_cols)
dicts = {}
for key, file in file_dict.items():
orig_dict = ColumnDict(skip_cols=skip_cols)
df = get_new_dataframe(file)
orig_dict.update(df)
dicts[key] = orig_dict
dicts_all.update_dict(orig_dict)
return dicts_all, dicts
def test_update_dict_with_value_cols(self):
stern_df_test1 = pd.read_csv(self.stern_test1_path,
delimiter='\t',
header=0)
stern_df_test3 = pd.read_csv(self.stern_test3_path,
delimiter='\t',
header=0)
dict1 = ColumnDict(value_cols=['latency'])
dict1.update(stern_df_test3)
dict2 = ColumnDict(value_cols=['latency'])
dict2.update(stern_df_test1)
dict2.update_dict(dict1)
dict1.update(stern_df_test1)
value_keys = dict1.value_info.keys()
for col in value_keys:
self.assertEqual(dict1.value_info[col], dict2.value_info[col],
"update_dict should update values correctly")