def generate_column_value_pairs(columns, n):
"""
function: dictionary, int -> generator
Takes a dictionary describing the table columns and generates test data
to be inserted into the table.
@return Return data in the form (but as a generator): [{"NAME": "Donald", "AGE": 54}, ...]
columns looks like: {"name": str, "age": int}
n corresponds to the number of rows of generated data to return
as a generator.
"""
# generates the distributions
gs = get_names()
gg = gaussian_generator(5, 5)
uuid = generate_uuid()
# return result
for _ in range(n):
data_dict = {}
for k, v in columns.items():
if v == dt.str:
data_dict[k] = next(gs)
elif v == dt.gaussian_int:
data_dict[k] = round(next(gg))
elif v == dt.uuid:
data_dict[k] = next(uuid)
else:
raise Exception("Type not found")
yield data_dict
def test_gaussian_generator(self):
""" Tests that a Gaussian distribution of numbers has the correct mean
and standard deviation."""
expected_mean = 10
expected_sigma = 15
gg = gaussian_generator(expected_mean, expected_sigma)
for _ in range(10):
distribution = [next(gg) for _ in range(10000)]
actual_mean = self.mean(distribution)
actual_sigma = self.standard_deviation(distribution)
# assert that the mean is about 10
self.assertGreater(actual_mean, 9.6)
self.assertLess(actual_mean, 10.4)
# assert that the standard deviation
# is around 15
self.assertGreater(actual_sigma, 14.5)
self.assertLess(actual_sigma, 15.5)
