def test_import_data(self):
"""Test the generalized import data function"""
# File test
data, var_names = utilities.import_data(self.data_path)
self.assertEqual(var_names, self.expected_var_names)
assert_array_equal(data, self.expected_arr)
# dict test
data_dict = utilities.csv_to_dict(self.data_path, dtype=float)
data_arr, var_names = utilities.import_data(data_dict)
self.assertEqual(var_names, self.expected_var_names)
assert_array_equal(data_arr, self.expected_arr)
# array test
data_arr_2, var_names = utilities.import_data(data_arr)
self.assertEqual(var_names, list(range(len(data_arr[0, :]))))
assert_array_equal(data_arr_2, self.expected_arr)
with self.assertRaises(NotImplementedError):
utilities.import_data([0, 1])
def test_no_header_row(self):
"""Test import with no header row"""
data = utilities.csv_to_dict(self.data_path_nh,
dtype=float,
header_row=False)
self.assertEqual(data, self.expected_dict_nh)
def test_csv_to_dict(self):
"""Test csv_to_dict"""
data = utilities.csv_to_dict(self.data_path, dtype=float)
self.assertEqual(data, self.expected_dict)
def test_dict_to_array(self):
"""Test dict_to_array"""
dict_data = utilities.csv_to_dict(self.data_path, dtype=float)
arr = utilities.dict_to_array(dict_data)
assert_array_equal(arr["data"], self.expected_arr)
self.assertEqual(arr["var_names"], self.expected_var_names)
def test_widen_data(self):
"""Test widen data script"""
data_dict = utilities.csv_to_dict(example_narrow_data)
kwargs = {
"uid_columns": ["patient", "plan", "field id"],
"x_data_cols": ["DD(%)", "DTA(mm)", "Threshold(%)"],
"y_data_col": "Gamma Pass Rate(%)",
"date_col": "date",
"dtype": float,
}
ds = utilities.widen_data(data_dict, **kwargs)
expected = {
"uid":
["ANON1234 && Plan_name && 3", "ANON1234 && Plan_name && 4"],
"date": ["6/13/2019 7:27", "6/13/2019 7:27"],
"2.0 && 3.0 && 10.0": [np.nan, 99.99476863],
"2.0 && 3.0 && 5.0": [99.88772435, 99.99533258],
"3.0 && 2.0 && 10.0": [99.94708217, 99.99941874],
"3.0 && 3.0 && 10.0": [99.97706894, 100],
"3.0 && 3.0 && 5.0": [99.97934552, 100],
}
for key, exp_value in expected.items():
assert_array_equal(ds[key], exp_value)
# No date test
kwargs_no_date = {key: value for key, value in kwargs.items()}
kwargs_no_date["date_col"] = None
ds_2 = utilities.widen_data(data_dict, **kwargs_no_date)
for key, ds_2_value in ds_2.items():
assert_array_equal(ds_2_value, expected[key])
# test column length check
data_dict_2 = utilities.csv_to_dict(example_narrow_data)
data_dict_2[list(data_dict_2)[0]].append("test")
with self.assertRaises(NotImplementedError):
utilities.widen_data(data_dict_2, **kwargs)
# test policy check
with self.assertRaises(NotImplementedError):
utilities.widen_data(data_dict, multi_val_policy="test", **kwargs)
ds = utilities.widen_data(data_dict, multi_val_policy="last", **kwargs)
index = ds["uid"].index("ANON1234 && Plan_name && 4")
self.assertEqual(ds["2.0 && 3.0 && 10.0"][index], 50)
ds = utilities.widen_data(data_dict, multi_val_policy="min", **kwargs)
index = ds["uid"].index("ANON1234 && Plan_name && 4")
self.assertEqual(ds["2.0 && 3.0 && 10.0"][index], 50)
ds = utilities.widen_data(data_dict, multi_val_policy="max", **kwargs)
index = ds["uid"].index("ANON1234 && Plan_name && 4")
self.assertEqual(
ds["2.0 && 3.0 && 10.0"][index],
expected["2.0 && 3.0 && 10.0"][index],
)
ds = utilities.widen_data(data_dict,
remove_partial_columns=True,
**kwargs)
self.assertTrue("2.0 && 3.0 && 10.0" not in ds.keys())