def test_sample_without_replacement():
array = np.ones(100)
node_single = DataNode()
node_single.set_private_data(name="array", data=array)
sample_size = 50
def u(x, r):
output = np.zeros(len(r))
for i in range(len(r)):
output[i] = r[i] * sum(np.greater_equal(x, r[i]))
return output
r = np.arange(0, 3.5, 0.001)
delta_u = r.max()
epsilon = 5
exponential_mechanism = ExponentialMechanism(u,
r,
delta_u,
epsilon,
size=sample_size)
access_modes = [LaplaceMechanism(1, 1)]
access_modes.append(GaussianMechanism(1, (0.5, 0.5)))
access_modes.append(RandomizedResponseBinary(0.5, 0.5, 1))
access_modes.append(RandomizedResponseCoins())
access_modes.append(exponential_mechanism)
for a in access_modes:
sampling_method = SampleWithoutReplacement(a, sample_size, array.shape)
node_single.configure_data_access("array", sampling_method)
result = node_single.query("array")
assert result.shape[0] == sample_size
def test_exponential_mechanism_pricing():
def u(x, r):
output = np.zeros(len(r))
for i in range(len(r)):
output[i] = r[i] * sum(np.greater_equal(x, r[i]))
return output
x = [1.00, 1.00, 1.00, 3.01] # Input dataset: the true bids
r = np.arange(0, 3.5, 0.001) # Set the interval of possible outputs r
delta_u = r.max() # In this specific case, Delta u = max(r)
epsilon = 5 # Set a value for epsilon
size = 10000 # We want to repeat the query this many times
node = DataNode()
node.set_private_data(name="bids", data=np.array(x))
data_access_definition = ExponentialMechanism(u, r, delta_u, epsilon, size)
node.configure_data_access("bids", data_access_definition)
result = node.query("bids")
y_bin, x_bin = np.histogram(a=result,
bins=int(round(np.sqrt(len(result)))),
density=True)
max_price = x_bin[np.where(y_bin == y_bin.max())]
min_price = x_bin[np.where(y_bin == y_bin.min())]
bin_size = x_bin[1] - x_bin[0]
assert (1.00 - x_bin[np.where(y_bin == max_price)] >
bin_size).all() # Check the best price is close to 1.00
assert ((x_bin[np.where(y_bin == min_price)] > (3.01 - bin_size)
).all() # Check the no-revenue price is either greater than 3.01
or x_bin[np.where(y_bin == min_price)][0] < bin_size
) # or close to 0.00
def test_configure_data_access():
data_access_definition = AdaptiveDifferentialPrivacy(epsilon_delta=(1, 1))
data_node = DataNode()
data_node.set_private_data("test", np.array(range(10)))
with pytest.raises(ValueError):
data_node.configure_data_access("test", data_access_definition)
data_node.query("test")
def test_exception_budget_2():
data_access_definition = AdaptiveDifferentialPrivacy(epsilon_delta=(1, 0.001))
data_node = DataNode()
array = np.array(range(10))
data_node.set_private_data("test", array)
data_node.configure_data_access("test", data_access_definition)
with pytest.raises(ExceededPrivacyBudgetError):
for i in range(1, 1000):
data_node.query("test", differentially_private_mechanism=GaussianMechanism(1, epsilon_delta=(0.1, 1)))
def test_randomize_binary_mechanism_no_binary_scalar():
scalar = 0.1
node_single = DataNode()
node_single.set_private_data(name="scalar", data=scalar)
data_access_definition = RandomizedResponseBinary(f0=0.5,
f1=0.5,
epsilon=1)
node_single.configure_data_access("scalar", data_access_definition)
with pytest.raises(ValueError):
node_single.query("scalar")
def test_randomize_binary_mechanism_scalar_coins():
scalar = 1
node_single = DataNode()
node_single.set_private_data(name="scalar", data=scalar)
node_single.configure_data_access("scalar", RandomizedResponseCoins())
result = node_single.query(private_property="scalar")
assert np.isscalar(result)
assert result == 0 or result == 1
def test_laplace_scalar_mechanism():
scalar = 175
node = DataNode()
node.set_private_data("scalar", scalar)
node.configure_data_access("scalar", LaplaceMechanism(1, 1))
result = node.query("scalar")
assert scalar != result
assert np.abs(scalar - result) < 100
def test_data_access():
data_access_definition = AdaptiveDifferentialPrivacy(epsilon_delta=(1, 1))
data_node = DataNode()
array = np.array(range(10))
data_node.set_private_data("test", array)
data_node.configure_data_access("test", data_access_definition)
query_result = data_node.query("test", differentially_private_mechanism=GaussianMechanism(1,
epsilon_delta=(0.1, 1)))
assert query_result is not None
def test_gaussian_scalar_mechanism():
scalar = 175
node = DataNode()
node.set_private_data("scalar", scalar)
node.configure_data_access("scalar",
GaussianMechanism(1, epsilon_delta=(0.1, 1)))
result = node.query("scalar")
assert scalar != result
assert np.abs(scalar - result) < 100
def test_exception_exceeded_privacy_budget_error():
scalar = 175
dp_mechanism = GaussianMechanism(1, epsilon_delta=(0.1, 1))
data_access_definition = AdaptiveDifferentialPrivacy(epsilon_delta=(1, 0),
differentially_private_mechanism=dp_mechanism)
node = DataNode()
node.set_private_data("scalar", scalar)
node.configure_data_access("scalar", data_access_definition)
with pytest.raises(ExceededPrivacyBudgetError):
node.query("scalar")
def test_sample_without_replacement_multidimensional():
array = np.ones((100, 2))
sample_size = 50
node_single = DataNode()
node_single.set_private_data(name="array", data=array)
epsilon = 1
access_mode = LaplaceMechanism(1, epsilon)
sampling_method = SampleWithoutReplacement(access_mode, sample_size,
array.shape)
node_single.configure_data_access("array", sampling_method)
result = node_single.query("array")
assert result.shape[0] == sample_size
def test_randomize_binary_mechanism_array_almost_always_true_values_coins():
array = np.ones(1000)
node_single = DataNode()
node_single.set_private_data(name="array", data=array)
# Very low heads probability in the first attempt, mean should be near true value
data_access_definition = RandomizedResponseCoins(prob_head_first=0.01,
prob_head_second=0.9)
node_single.configure_data_access("array", data_access_definition)
result = node_single.query("array")
assert 1 - np.mean(result) < 0.05
def test_randomize_binary_random_scalar_0():
scalar = 0
node_single = DataNode()
node_single.set_private_data(name="scalar", data=scalar)
data_access_definition = RandomizedResponseBinary(f0=0.5,
f1=0.5,
epsilon=1)
node_single.configure_data_access("scalar", data_access_definition)
result = node_single.query(private_property="scalar")
assert np.isscalar(result)
assert result == 0 or result == 1
def test_laplace_dictionary_mechanism():
dictionary = {
0: np.array([[2, 4, 5], [2, 3, 5]]),
1: np.array([[1, 3, 1], [1, 4, 6]])
}
node = DataNode()
node.set_private_data("dictionary", dictionary)
node.configure_data_access("dictionary", LaplaceMechanism(1, 1))
result = node.query("dictionary")
assert dictionary.keys() == result.keys()
assert np.mean(dictionary[0]) - np.mean(result[0]) < 5
def test_randomize_binary_mechanism_array_almost_always_false_values_zeros():
array = np.zeros(1000)
node_single = DataNode()
node_single.set_private_data(name="array", data=array)
# Prob of one given 1 very low, mean should be near 0
data_access_definition = RandomizedResponseBinary(f0=0.01,
f1=0.5,
epsilon=1)
node_single.configure_data_access("array", data_access_definition)
result = node_single.query("array")
assert np.abs(1 - np.mean(result)) < 0.05
def test_randomize_binary_mechanism_array_almost_always_true_values_ones():
array = np.ones(1000)
node_single = DataNode()
node_single.set_private_data(name="array", data=array)
# Prob of one given 1 very high, mean should be near 1
data_access_definition = RandomizedResponseBinary(f0=0.5,
f1=0.99,
epsilon=5)
node_single.configure_data_access("array", data_access_definition)
result = node_single.query("array")
assert 1 - np.mean(result) < 0.05
def test_laplace_dictionary_mechanism_wrong_shapes():
dictionary = {0: np.array([2, 3, 5]), 1: np.array([[1, 3, 1], [1, 4, 6]])}
sensitivity = {
0: np.array([[1, 1, 2], [2, 1, 1]]),
1: np.array([3, 1, 11, 1, 2])
}
node = DataNode()
node.set_private_data("dictionary", dictionary)
dp_access_mechanism = LaplaceMechanism(sensitivity, 1)
node.configure_data_access("dictionary", dp_access_mechanism)
with pytest.raises(ValueError):
node.query("dictionary")
def test_randomize_binary_mechanism_array_coins():
array = np.ones(100)
node_single = DataNode()
node_single.set_private_data(name="array", data=array)
node_single.configure_data_access("array", RandomizedResponseCoins())
result = node_single.query("array")
differences = 0
for i in range(100):
if result[i] != array[i]:
differences = differences + 1
assert not np.isscalar(result)
assert 0 < differences < 100
assert np.mean(result) < 1
def test_randomize_binary_random():
data_size = 100
array = np.ones(data_size)
node_single = DataNode()
node_single.set_private_data(name="A", data=array)
data_access_definition = RandomizedResponseBinary(f0=0.5,
f1=0.5,
epsilon=1)
node_single.configure_data_access("A", data_access_definition)
result = node_single.query(private_property="A")
differences = 0
for i in range(data_size):
if result[i] != array[i]:
differences = differences + 1
assert 0 < differences < data_size
assert np.mean(result) < 1
def test_exponential_mechanism_obtain_laplace():
def u_laplacian(x, r):
output = -np.absolute(x - r)
return output
r = np.arange(-20, 20, 0.001) # Set the interval of possible outputs r
x = 3.5 # Set a value for the dataset
delta_u = 1 # We simply set it to one
epsilon = 1 # Set a value for epsilon
size = 100000 # We want to repeat the query this many times
node = DataNode()
node.set_private_data(name="identity", data=np.array(x))
data_access_definition = ExponentialMechanism(u_laplacian, r, delta_u,
epsilon, size)
node.configure_data_access("identity", data_access_definition)
result = node.query("identity")
assert (result > r.min()).all() and (
result < r.max()).all() # Check all outputs are within range
assert np.absolute(np.mean(result) - x) < (
delta_u / epsilon) # Check the mean output is close to true value
def test_sensitivity_wrong_input():
epsilon_delta = (0.1, 1)
# Negative sensitivity:
scalar = 175
sensitivity = -0.1
node = DataNode()
node.set_private_data("scalar", scalar)
with pytest.raises(ValueError):
node.configure_data_access(
"scalar",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
# Scalar query result, Too many sensitivity values provided:
scalar = 175
sensitivity = [0.1, 0.5]
node = DataNode()
node.set_private_data("scalar", scalar)
node.configure_data_access(
"scalar",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("scalar")
# Both query result and sensitivity are 1D-arrays, but non-broadcastable:
data_array = [10, 10, 10, 10]
sensitivity = [0.1, 10, 100, 1000, 1000]
node = DataNode()
node.set_private_data("data_array", data_array)
node.configure_data_access(
"data_array",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("data_array")
# ND-array query result and 1D-array sensitivity, but non-broadcastable:
data_ndarray = [[10, 10, 10, 10], [10, 10, 10, 10], [10, 10, 10, 10]]
sensitivity = [0.1, 10, 100]
node = DataNode()
node.set_private_data("data_ndarray", data_ndarray)
node.configure_data_access(
"data_ndarray",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("data_ndarray")
# Both query result and sensitivity are ND-arrays, but non-broadcastable (they should have the same shape in this case):
data_ndarray = [[10, 10, 10, 10], [10, 10, 10, 10], [10, 10, 10, 10]]
sensitivity = [[0.1, 10, 100, 1000, 10000], [0.1, 10, 100, 1000, 10000],
[0.1, 10, 100, 1000, 10000]]
node = DataNode()
node.set_private_data("data_ndarray", data_ndarray)
node.configure_data_access(
"data_ndarray",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("data_ndarray")
def test_sensitivity_wrong_input():
epsilon_delta = (0.1, 1)
# Negative sensitivity:
scalar = 175
sensitivity = -0.1
node = DataNode()
node.set_private_data("scalar", scalar)
with pytest.raises(ValueError):
node.configure_data_access(
"scalar",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
# Scalar query result, Too many sensitivity values provided:
scalar = 175
sensitivity = [0.1, 0.5]
node = DataNode()
node.set_private_data("scalar", scalar)
node.configure_data_access(
"scalar",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("scalar")
# Both query result and sensitivity are 1D-arrays, but non-broadcastable:
data_array = [10, 10, 10, 10]
sensitivity = [0.1, 10, 100, 1000, 1000]
node = DataNode()
node.set_private_data("data_array", data_array)
node.configure_data_access(
"data_array",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("data_array")
# ND-array query result and 1D-array sensitivity, but non-broadcastable:
data_ndarray = [[10, 10, 10, 10], [10, 10, 10, 10], [10, 10, 10, 10]]
sensitivity = [0.1, 10, 100]
node = DataNode()
node.set_private_data("data_ndarray", data_ndarray)
node.configure_data_access(
"data_ndarray",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("data_ndarray")
# Both query result and sensitivity are ND-arrays, but non-broadcastable (they should have the same shape
# in this case):
data_ndarray = [[10, 10, 10, 10], [10, 10, 10, 10], [10, 10, 10, 10]]
sensitivity = [[0.1, 10, 100, 1000, 10000], [0.1, 10, 100, 1000, 10000],
[0.1, 10, 100, 1000, 10000]]
node = DataNode()
node.set_private_data("data_ndarray", data_ndarray)
node.configure_data_access(
"data_ndarray",
GaussianMechanism(sensitivity=sensitivity,
epsilon_delta=epsilon_delta))
with pytest.raises(ValueError):
result = node.query("data_ndarray")
# Query result is a list of arrays: sensitivity must be either a scalar, or a list of the same length as query
data_list = [
np.random.rand(30, 20),
np.random.rand(20, 30),
np.random.rand(50, 40)
]
sensitivity = np.array([1, 1]) # Array instead of scalar
node = DataNode()
node.set_private_data("data_list", data_list)
node.configure_data_access(
"data_list", LaplaceMechanism(sensitivity=sensitivity, epsilon=1))
with pytest.raises(ValueError):
result = node.query("data_list")
sensitivity = [1, 1] # List of wrong length
node = DataNode()
node.set_private_data("data_list", data_ndarray)
node.configure_data_access(
"data_list", LaplaceMechanism(sensitivity=sensitivity, epsilon=1))
with pytest.raises(IndexError):
result = node.query("data_list")
# Query result is wrong data structure: so far, tuples are not allowed
data_tuple = (1, 2, 3, 4, 5)
sensitivity = 2
node = DataNode()
node.set_private_data("data_tuple", data_tuple)
node.configure_data_access(
"data_tuple", LaplaceMechanism(sensitivity=sensitivity, epsilon=1))
with pytest.raises(NotImplementedError):
result = node.query("data_tuple")