def test_add_raised_suppressed_error(self):
"""Test that an supressed error does not show up in the unsuppressed error list."""
report = ValidationReport([self.ERROR_1])
report.add_raised_error(self.ERROR_1)
assert len(report.unsuppressed_errors) == 0
assert report.num_suppressed == 1
def test_add_raised_unsuppressed_error(self):
"""Test that an unsupressed error shows up in the unsuppressed error list."""
report = ValidationReport([ValidationFailure("good",
filename="20201107_county_sig2.csv",
message="msg 2")])
report.add_raised_error(self.ERROR_1)
report.add_raised_error(self.ERROR_2)
assert report.unsuppressed_errors == [self.ERROR_1, self.ERROR_2]
def test_str(self):
"""Test that the string representation contains all information."""
report = ValidationReport([self.ERROR_1])
report.increment_total_checks()
report.increment_total_checks()
report.increment_total_checks()
report.add_raised_warning(ImportWarning("wrong import"))
report.add_raised_warning(ImportWarning("right import"))
report.add_raised_error(self.ERROR_1)
report.add_raised_error(self.ERROR_2)
def test_no_outlier(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
# Data from 51580 between 9/24 and 10/26 (10/25 query date)
ref_val = [30, 30.28571429, 30.57142857, 30.85714286, 31.14285714,
31.42857143, 31.71428571, 32, 32, 32.14285714,
32.28571429, 32.42857143, 32.57142857, 32.71428571,
32.85714286, 33, 33, 33, 33, 33, 33, 33, 33,
33, 33, 33, 33, 33, 33, 33]
test_val = [33, 33, 33]
ref_data = {"val": ref_val, "se": [np.nan] * len(ref_val),
"sample_size": [np.nan] * len(ref_val), "geo_id": ["1"] * len(ref_val),
"time_value": pd.date_range(start="2020-09-24", end="2020-10-23")}
test_data = {"val": test_val, "se": [np.nan] * len(test_val),
"sample_size": [np.nan] * len(test_val), "geo_id": ["1"] * len(test_val),
"time_value": pd.date_range(start="2020-10-24", end="2020-10-26")}
ref_data2 = {"val": ref_val, "se": [np.nan] * len(ref_val),
"sample_size": [np.nan] * len(ref_val), "geo_id": ["2"] * len(ref_val),
"time_value": pd.date_range(start="2020-09-24", end="2020-10-23")}
test_data2 = {"val": test_val, "se": [np.nan] * len(test_val),
"sample_size": [np.nan] * len(test_val), "geo_id": ["2"] * len(test_val),
"time_value": pd.date_range(start="2020-10-24", end="2020-10-26")}
ref_df = pd.concat([pd.DataFrame(ref_data), pd.DataFrame(ref_data2)]). \
reset_index(drop=True)
test_df = pd.concat([pd.DataFrame(test_data), pd.DataFrame(test_data2)]). \
reset_index(drop=True)
validator.check_positive_negative_spikes(
test_df, ref_df, "state", "signal", report)
assert len(report.raised_errors) == 0
def test_more_than_two_copies(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame([["a", "1"], ["b", "2"], ["b", "2"], ["b", "2"]])
validator.check_duplicate_rows(df, FILENAME, report)
assert len(report.raised_warnings) == 1
assert report.raised_warnings[0].check_name == "check_duplicate_rows"
def test_1000x_val(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
test_data = {
"val": [1, 1, 1, 2000, 0, 1],
"se": [np.nan] * 6,
"sample_size": [np.nan] * 6,
"geo_id": ["1"] * 6
}
ref_data = {
"val": [1, 1, 1, 2, 0, 1],
"se": [np.nan] * 6,
"sample_size": [np.nan] * 6,
"geo_id": ["1"] * 6
}
test_df = pd.DataFrame(test_data)
ref_df = pd.DataFrame(ref_data)
validator.check_avg_val_vs_reference(
test_df, ref_df, datetime.combine(date.today(),
datetime.min.time()), "geo",
"signal", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[
0].check_name == "check_test_vs_reference_avg_changed"
def test_empty_df(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
empty_df = pd.DataFrame(columns=["geo_id"], dtype=str)
validator.check_bad_geo_id_format(empty_df, FILENAME, "county", report)
assert len(report.raised_errors) == 0
def mock_validator_fn():
"""Set up a mock validator function."""
validator_fn = mock.Mock()
validator = mock.Mock()
validator_fn.return_value = validator
validator.validate.return_value = ValidationReport([])
yield validator_fn
def test_gt_max_prop(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame([1e7], columns=["val"])
validator.check_bad_val(df, FILENAME, "prop", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_val_prop_gt_100k"
def test_lt_0(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame([-5], columns=["val"])
validator.check_bad_val(df, FILENAME, "signal", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_val_lt_0"
def test_invalid_geo_id_value_nation(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame(["us", "zz"], columns=["geo_id"])
validator.check_bad_geo_id_value(df, FILENAME, "nation", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_bad_geo_id_value"
def test_invalid_geo_id_format_hhs(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame(["1", "112"], columns=["geo_id"])
validator.check_bad_geo_id_format(df, FILENAME, "hhs", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_geo_id_format"
def test_invalid_geo_type(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
empty_df = pd.DataFrame(columns=["geo_id"], dtype=str)
validator.check_bad_geo_id_format(empty_df, FILENAME, "hello", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_geo_type"
def test_empty_filelist(self):
params = {
"common": {
"data_source": "",
"span_length": 8,
"end_date": "2020-09-09"
}
}
validator = StaticValidator(params)
report = ValidationReport([])
report = ValidationReport([])
filenames = list()
validator.check_missing_date_files(filenames, report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_missing_date_files"
def test_empty_df(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
empty_df = pd.DataFrame(columns=["val"])
validator.check_bad_val(empty_df, "", "", report)
validator.check_bad_val(empty_df, "", "prop", report)
validator.check_bad_val(empty_df, "", "pct", report)
assert len(report.raised_errors) == 0
def test_same_df(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
test_df = pd.DataFrame([date.today()] * 5, columns=["time_value"])
ref_df = pd.DataFrame([date.today()] * 5, columns=["time_value"])
validator.check_rapid_change_num_rows(test_df, ref_df, date.today(),
"geo", "signal", report)
assert len(report.raised_errors) == 0
def test_uppercase_geo_id(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame(["ak", "AK"], columns=["geo_id"])
validator.check_bad_geo_id_value(df, FILENAME, "state", report)
assert len(report.raised_errors) == 0
assert len(report.raised_warnings) == 1
assert report.raised_warnings[0].check_name == "check_geo_id_lowercase"
def test_invalid_geo_id_format_state(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame(["aa", "hi", "HI", "hawaii", "Hawaii", "a", "H.I."],
columns=["geo_id"])
validator.check_bad_geo_id_format(df, FILENAME, "state", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_geo_id_format"
def test_lt_min_missing_not_allowed(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
validator.params.missing_sample_size_allowed = False
df = pd.DataFrame([[1, 0, 10], [1, np.nan, 240], [1, np.nan, 245]],
columns=["val", "se", "sample_size"])
validator.check_bad_sample_size(df, FILENAME, report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_n_gt_min"
def test_neg_outlier(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
ref_val = [
100, 101, 100, 101, 100, 100, 100, 100, 100, 100, 100, 102, 100,
100, 100, 100, 100, 101, 100, 100, 100, 100, 100, 99, 100, 100, 98,
100, 100, 100
]
test_val = [10, 10, 10]
ref_data = {
"val": ref_val,
"se": [np.nan] * len(ref_val),
"sample_size": [np.nan] * len(ref_val),
"geo_id": ["1"] * len(ref_val),
"time_value": pd.date_range(start="2020-09-24", end="2020-10-23")
}
test_data = {
"val": test_val,
"se": [np.nan] * len(test_val),
"sample_size": [np.nan] * len(test_val),
"geo_id": ["1"] * len(test_val),
"time_value": pd.date_range(start="2020-10-24", end="2020-10-26")
}
ref_data2 = {
"val": ref_val,
"se": [np.nan] * len(ref_val),
"sample_size": [np.nan] * len(ref_val),
"geo_id": ["2"] * len(ref_val),
"time_value": pd.date_range(start="2020-09-24", end="2020-10-23")
}
test_data2 = {
"val": test_val,
"se": [np.nan] * len(test_val),
"sample_size": [np.nan] * len(test_val),
"geo_id": ["2"] * len(test_val),
"time_value": pd.date_range(start="2020-10-24", end="2020-10-26")
}
ref_df = pd.concat([pd.DataFrame(ref_data), pd.DataFrame(ref_data2)]). \
reset_index(drop=True)
test_df = pd.concat([pd.DataFrame(test_data), pd.DataFrame(test_data2)]). \
reset_index(drop=True)
validator.check_positive_negative_spikes(test_df, ref_df, "state",
"signal", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[
0].check_name == "check_positive_negative_spikes"
def test_0_vs_many(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
time_value = datetime.combine(date.today(), datetime.min.time())
test_df = pd.DataFrame([time_value] * 5, columns=["time_value"])
ref_df = pd.DataFrame([time_value] * 1, columns=["time_value"])
validator.check_rapid_change_num_rows(
test_df, ref_df, time_value, "geo", "signal", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_rapid_change_num_rows"
def test_e_0_missing_not_allowed(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
validator.params.missing_se_allowed = False
df = pd.DataFrame([[1, 0, 200], [1, 0, np.nan], [1, np.nan, np.nan]],
columns=["val", "se", "sample_size"])
validator.check_bad_se(df, FILENAME, report)
assert len(report.raised_errors) == 2
assert "check_se_not_missing_and_in_range" in [
err.check_name for err in report.raised_errors
]
assert "check_se_0" in [err.check_name for err in report.raised_errors]
def test_state_level_fips(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
df = pd.DataFrame(["37183", "56000", "04000", "60000", "78000"], columns=["geo_id"])
validator.check_bad_geo_id_value(df, FILENAME, "county", report)
assert len(report.raised_errors) == 0
df = pd.DataFrame(["37183", "56000", "04000", "60000", "78000", "99000"], columns=["geo_id"])
validator.check_bad_geo_id_value(df, FILENAME, "county", report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_bad_geo_id_value"
def test_empty_df(self):
validator = StaticValidator(self.params)
report = ValidationReport([])
empty_df = pd.DataFrame(columns=["val", "se", "sample_size"],
dtype=float)
validator.check_bad_sample_size(empty_df, "", report)
assert len(report.raised_errors) == 0
validator.params.missing_sample_size_allowed = True
validator.check_bad_sample_size(empty_df, "", report)
assert len(report.raised_errors) == 0
def test_no_padding(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
data = {"val": [1, 1, 1, 2, 0, 1], "se": [np.nan] * 6,
"sample_size": [np.nan] * 6, "geo_id": ["1"] * 6,
"time_value": pd.date_range(start="2021-01-01", end="2021-01-06")}
test_df = pd.DataFrame(data)
ref_df = pd.DataFrame(data)
new_ref_df = validator.pad_reference_api_df(
ref_df, test_df, datetime.strptime("2021-01-06", "%Y-%m-%d").date())
assert new_ref_df.equals(ref_df)
def test_same_day(self):
params = {
"common": {
"data_source": "",
"span_length": 0,
"end_date": "2020-09-01"
}
}
validator = StaticValidator(params)
report = ValidationReport([])
filenames = [("20200901_county_signal_signal.csv", "match_obj")]
validator.check_missing_date_files(filenames, report)
assert len(report.raised_errors) == 0
def test_same_val_se_n(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
data = {"val": [1, 1, 1, 2, 0, 1, 1]*2, "se": [1, 1, 1, 2, 0, 1, 1]*2,
"sample_size": [1, 1, 1, 2, 0, 1, 1]*2, "geo_id": ["1"] * 14,
"time_value": ["2021-01-01", "2021-01-02", "2021-01-03", "2021-01-04", "2021-01-05", "2021-01-06", "2021-01-07",
"2021-01-08", "2021-01-09", "2021-01-10", "2021-01-11", "2021-01-12", "2021-01-13", "2021-01-14"]}
test_df = pd.DataFrame(data)
ref_df = pd.DataFrame(data)
validator.check_avg_val_vs_reference(
test_df, ref_df, date.today(), "geo", "signal", report)
assert len(report.raised_errors) == 0
def test_same_n(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
data = {
"val": [np.nan] * 6,
"se": [np.nan] * 6,
"sample_size": [1, 1, 1, 2, 0, 1],
"geo_id": ["1"] * 6
}
test_df = pd.DataFrame(data)
ref_df = pd.DataFrame(data)
validator.check_avg_val_vs_reference(test_df, ref_df, date.today(),
"geo", "signal", report)
assert len(report.raised_errors) == 0
def test_duplicate_dates(self):
params = {
"common": {
"data_source": "",
"span_length": 1,
"end_date": "2020-09-02"
}
}
validator = StaticValidator(params)
report = ValidationReport([])
filenames = [("20200901_county_signal_signal.csv", "match_obj"),
("20200903_county_signal_signal.csv", "match_obj"),
("20200903_usa_signal_signal.csv", "match_obj"),
("20200903_usa_signal_signal.csv", "match_obj")]
validator.check_missing_date_files(filenames, report)
assert len(report.raised_errors) == 1
assert report.raised_errors[0].check_name == "check_missing_date_files"
def test_half_padding(self):
validator = DynamicValidator(self.params)
report = ValidationReport([])
ref_data = {"val": [2, 2, 2, 2, 2, 2], "se": [np.nan] * 6,
"sample_size": [np.nan] * 6, "geo_id": ["1"] * 6,
"time_value": pd.date_range(start="2021-01-01", end="2021-01-06")}
test_data = {"val": [1, 1, 1, 1, 1, 1], "se": [np.nan] * 6,
"sample_size": [np.nan] * 6, "geo_id": ["1"] * 6,
"time_value": pd.date_range(start="2021-01-06", end="2021-01-11")}
ref_df = pd.DataFrame(ref_data)
test_df = pd.DataFrame(test_data)
new_ref_df = validator.pad_reference_api_df(
ref_df, test_df, datetime.strptime("2021-01-15", "%Y-%m-%d").date())
# Check it only takes missing dates - so the last 5 dates
assert new_ref_df.time_value.max() == datetime.strptime("2021-01-11",
"%Y-%m-%d").date()
assert new_ref_df.shape[0] == 11
assert new_ref_df.loc[:, "val"].iloc[5] == 2
