def test_check_int(self):
"""
Checks if number is integer.
:return:
"""
true_asserts = [
1,
1345,
-13,
0,
-0, # numeric values
"1" # strings
]
for assert_val in true_asserts:
self.assertTrue(NumericStatsMixin.is_int(assert_val))
false_asserts = [
1.3, # float
float("nan"),
np.nan, # nan value
"nan",
"1a",
"abc",
"",
"1.3" # strings
]
for assert_val in false_asserts:
self.assertFalse(NumericStatsMixin.is_int(assert_val))
def test_base(self):
# validate requires NumericalOptions
with self.assertRaisesRegex(
ValueError, "NumericalStatsMixin parameter 'options' "
"must be of type NumericalOptions."):
profile = NumericStatsMixin(options='bad options')
try:
# validate doesn't fail
profile = NumericStatsMixin()
profile = NumericStatsMixin(NumericalOptions())
except Exception as e:
self.fail(e)
def test_get_percentile_median(self):
num_profiler = TestColumn()
# Dummy data for calculating bin error
num_profiler._stored_histogram = {
"histogram": {
"bin_counts": np.array([1, 2, 0, 2, 1]),
"bin_edges": np.array([0.0, 4.0, 8.0, 12.0, 16.0, 20.0]),
}
}
median = NumericStatsMixin._get_percentile(num_profiler,
percentiles=[50, 50])
self.assertListEqual([10, 10], median)
def test_report(self):
options = NumericalOptions()
options.max.is_enabled = False
options.min.is_enabled = False
options.histogram_and_quantiles.is_enabled = False
options.variance.is_enabled = False
num_profiler = NumericStatsMixin(options=options)
num_profiler.match_count = 0
num_profiler.times = defaultdict(float)
report = num_profiler.report(remove_disabled_flag=True)
report_keys = list(report.keys())
for disabled_key in [
"max", "min", "variance", "histogram", "quantiles"
]:
self.assertNotIn(disabled_key, report_keys)
# test report default `remove_disabled_flag`
# value and no NumericalOptions
report = num_profiler.report()
report_keys = list(report.keys())
for disabled_key in [
"max", "min", "variance", "histogram", "quantiles"
]:
self.assertIn(disabled_key, report_keys)
def test_check_float(self):
"""
Checks if number is float.
:return:
"""
true_asserts = [
1.3,
1.345,
-1.3,
0.03,
0.0,
-0.0,
1, # numeric values
float("nan"),
np.nan, # nan values
"1.3",
"nan" # strings
]
for assert_val in true_asserts:
self.assertTrue(NumericStatsMixin.is_float(assert_val))
false_asserts = ["1.3a", "abc", "", "1.23.45"]
for assert_val in false_asserts:
self.assertFalse(NumericStatsMixin.is_float(assert_val))
def __init__(self):
NumericStatsMixin.__init__(self)
self.match_count = 0
self.times = defaultdict(float)