def test_ScaledArray_array():
N = 50
P = 40
array = np.random.rand(N, P) + 1
std = np.diag(1 / np.std(array, axis=0))
mu = np.mean(array, axis=0)
for s in [True, False]:
for c in [True, False]:
for f in [None, 'n', 'p']:
sarray = ScaledCenterArray(scale=s, center=c, factor=f)
sarray.fit(da.array(array))
if c:
array = array - mu
if s:
array = array.dot(std)
result = array.dot(array.T)
if f == 'n':
result /= N
elif f == 'p':
result /= P
np.testing.assert_array_almost_equal(sarray.array, result)
def test_array_shapes():
N, K = 10, 7
array = da.array(np.random.rand(N, K))
sarray = ScaledCenterArray()
sarray.fit(array)
assert sarray.shape == (N, K)
assert sarray.T.shape == (K, N)
def test__getitem__base_array_case2():
for N in range(2, 5):
for P in range(2, 5):
array = da.array(np.random.rand(N , P))
sarray = ScaledCenterArray()
sarray.fit(array)
for i in range(2, N):
for j in range(2, P):
np.testing.assert_array_equal(sarray[0:i, 0:j]._array, array[0:i, 0:j])
def test_array_id():
array = da.array(np.random.rand(10, 7))
x = da.array(np.random.rand(10,5))
sarray = ScaledCenterArray(scale=True, center=True)
sarray.fit(da.array(array), x=x)
sarray_T = sarray.T
assert id(sarray._array) == id(sarray_T._array)
assert id(sarray.center_vector) == id(sarray_T.center_vector)
assert id(sarray._array_moment.scale_matrix) == id(sarray_T._array_moment.scale_matrix)
assert id(sarray._array_moment.sym_scale_matrix) == id(sarray_T._array_moment.sym_scale_matrix)
def test_array_tranpose_tranpose():
array = da.array(np.random.rand(7, 10))
x = da.array(np.random.rand(10, 5))
sarray = ScaledCenterArray(scale=True, center=True)
sarray.fit(da.array(array))
s_array_T_T = sarray.T.T
assert id(sarray._array) == id(s_array_T_T._array)
assert id(sarray) == id(s_array_T_T)
np.testing.assert_array_equal(sarray.dot(x), s_array_T_T.dot(x))
def test_ScaledArray_fromArrayMoment_moments():
N1, P = 7, 10
N2 = 5
array1 = da.random.random(size=(N1, P)).persist()
array2 = da.random.random(size=(N2, P)).persist()
sa1 = ScaledCenterArray(scale=True, center=True, factor='n')
sa1.fit(array1)
sa2 = ScaledCenterArray.fromScaledArray(array=array2, scaled_array=sa1, factor='n')
np.testing.assert_array_equal(sa2.center_vector, sa1.center_vector)
np.testing.assert_array_equal(sa2.scale_vector, sa1.scale_vector)
assert sa2.factor_value == N2
assert sa1.factor_value == N1
def test__getitem_T_subset_chunks():
a = da.random.random(size=(100, 200))
sa = ScaledCenterArray()
sa.fit(a)
np.testing.assert_array_equal(sa.T[0:10, :].chunks, a.T[0:10, :].chunks)
np.testing.assert_array_equal(sa.T[:, 0:8].chunks, a.T[:, 0:8].chunks)
np.testing.assert_array_equal(sa.T[0:10, 0:8].chunks, a.T[0:10, 0:8].chunks)
np.testing.assert_array_equal(sa[0:10, :].chunks, a[0:10, :].chunks)
np.testing.assert_array_equal(sa.T[:, 0:8].chunks, a.T[:, 0:8].chunks)
np.testing.assert_array_equal(sa.T[0:10, 0:8].chunks, a.T[0:10, 0:8].chunks)
np.testing.assert_array_equal(sa[0:50, 0:40].T[0:20, 0:30].chunks, a[0:50, 0:40].T[0:20, 0:30].chunks)
def test__getitem_scaled_array_shape():
for N in range(2, 5):
for P in range(2, 5):
array = da.array(np.random.rand(N, P))
sarray = ScaledCenterArray()
sarray.fit(array)
for i in range(2, N):
for j in range(2, P):
assert sarray[0:i, 0:j].shape == array[0:i, 0:j].shape
for size in range(2, 5):
i = np.sort(np.random.choice(np.arange(N), size=size))
assert sarray[i, :].shape == array[i, :].shape
def test_ScaledArray_std_method():
for method in [None, 'normal']:
array = np.random.randint(0, 3, size=(10, 20))
sa = ScaledCenterArray(std_dist=method)
sa.fit(array)
np.testing.assert_almost_equal(sa.scale_vector, 1/array.std(axis=0))
sa = ScaledCenterArray(std_dist='binom')
sa.fit(array)
p = array.mean(axis=0) / 2
np.testing.assert_almost_equal(sa.scale_vector, 1/np.sqrt(2*p*(1-p)))
def test_ScaledArray_fromArrayMoment_array():
N1, P = 7, 10
N2 = 5
array1 = da.random.random(size=(N1, P)).persist()
mu = da.mean(array1, axis=0)
std = da.diag(1/da.std(array1, axis=0))
array2 = da.random.random(size=(N2, P)).persist()
for scale in [True, False]:
for center in [True, False]:
for factor1 in [None, 'n', 'p']:
sa1 = ScaledCenterArray(scale=scale, center=center, factor=factor1)
sa1.fit(array1)
for factor2, factor_value in zip([None, 'n', 'p'], [1, N2, P]):
sa2 = ScaledCenterArray.fromScaledArray(array=array2, scaled_array=sa1, factor=factor2)
sa2_array = array2
if center:
sa2_array = sa2_array - mu
if scale:
sa2_array = sa2_array.dot(std)
np.testing.assert_array_almost_equal(sa2.array, sa2_array)
def test_ScaledArray_fromArrayMoment_bad_array():
N1, P1 = 7, 10
N2, P2 = 7, 9
array1 = da.random.random(size=(N1, P1)).persist()
array2 = da.random.random(size=(N2, P2)).persist()
sa1 = ScaledCenterArray(scale=True, center=True, factor='n')
sa1.fit(array1)
with pytest.raises(ValueError):
ScaledCenterArray.fromScaledArray(array=array2, scaled_array=sa1, factor='n')
def test__getitem__mult_case():
for N in range(3, 5):
for P in range(3, 5):
array = np.random.rand(N, P) + 1
for K in range(1, 5):
for sub_N in range(2, N):
sub_array = array[0:sub_N, :]
std = np.diag(1 / np.std(sub_array, axis=0))
mu = np.mean(sub_array, axis=0)
for squeeze in [True, False]:
x = np.random.rand(P, K)
if squeeze:
x = np.squeeze(x)
for fit_x in [x, None]:
# With No Scale or Center
# x = A'Ax
result = sub_array.dot(x)
sarray = ScaledCenterArray(scale=False, center=False)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray[0:sub_N, :].dot(x))
# With Scale but No Center
# B = AD
b_array = sub_array.dot(std)
result = b_array.dot(x)
sarray = ScaledCenterArray(scale=True, center=False)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray[0:sub_N, :].dot(x))
# With Center but No Scale:
# B = (A - U)
b_array = sub_array - mu
result = b_array.dot(x)
sarray = ScaledCenterArray(scale=False, center=True)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray[0:sub_N, :].dot(x))
# With Center and Scale:
# (A - U)'D'D(A - U)x
b_array = (sub_array - mu).dot(std)
result = b_array.dot(x)
sarray = ScaledCenterArray(scale=True, center=True)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray[0:sub_N, :].dot(x))
def test_ScaledArray_array():
for N in range(3, 5):
for P in range(3, 5):
array = np.random.rand(N, P) + 1
std = np.diag(1 / np.std(array, axis=0))
mu = np.mean(array, axis=0)
sarray = ScaledCenterArray(scale=False, center=False)
sarray.fit(da.array(array))
np.testing.assert_array_almost_equal(array, sarray.array)
np.testing.assert_array_almost_equal(array.T, sarray.T.array)
# With Scale but No Center
# B = AD
b_array = array.dot(std)
sarray = ScaledCenterArray(scale=True, center=False)
sarray.fit(da.array(array))
np.testing.assert_array_almost_equal(b_array, sarray.array)
np.testing.assert_array_almost_equal(b_array.T, sarray.T.array)
# With Center but No Scale:
# B = (A - U)
b_array = array - mu
sarray = ScaledCenterArray(scale=False, center=True)
sarray.fit(da.array(array))
np.testing.assert_array_almost_equal(b_array, sarray.array)
np.testing.assert_array_almost_equal(b_array.T, sarray.T.array)
# With Center and Scale:
# (A - U)'D'D(A - U)x
b_array = (array - mu).dot(std)
sarray = ScaledCenterArray(scale=True, center=True)
sarray.fit(da.array(array))
np.testing.assert_array_almost_equal(b_array, sarray.array)
np.testing.assert_array_almost_equal(b_array.T, sarray.T.array)
def test__getitem__base_array_case1():
array = da.array(np.random.rand(3, 3))
sarray = ScaledCenterArray()
sarray.fit(array)
np.testing.assert_array_equal(sarray[:, :]._array, sarray._array)
def test_bad_x():
array = da.array(np.random.rand(3, 3))
x = da.array(np.random.rand(3,3,3))
sarray = ScaledCenterArray()
with pytest.raises(ValueError):
sarray.fit(array, x)
def test_ScaledArray_fromArrayMoment_not_fit_array():
N1, P1 = 7, 10
array2 = da.random.random(size=(N1, P1)).persist()
sa1 = ScaledCenterArray(scale=True, center=True, factor='n')
with pytest.raises(AttributeError):
ScaledCenterArray.fromScaledArray(array=array2, scaled_array=sa1, factor='n')
def test_ScaledArray_sym_mat_mult():
for N in range(2, 5):
for P in range(2, 5):
array = np.random.rand(N, P) + 1
std = np.diag(1/np.std(array, axis=0))
mu = np.mean(array, axis=0)
for factor in [None, 'n', 'p']:
if factor is None:
f = 1
elif factor == 'n':
f = N
else:
f = P
for K in range(1, 5):
for squeeze in [True, False]:
x = np.random.rand(N, K)
if squeeze:
x = np.squeeze(x)
for fit_x in [x, None]:
# With No Scale or Center
# x = A'Ax
result = array.dot(array.T.dot(x))/f
assert result.shape == x.shape
sarray = ScaledCenterArray(scale=False, center=False, factor=factor)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_equal(result, sarray.sym_mat_mult(x))
# With Scale but No Center
# B = AD
b_array = array.dot(std)
result = b_array.dot(b_array.T.dot(x))/f
assert result.shape == x.shape
sarray = ScaledCenterArray(scale=True, center=False, factor=factor)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.sym_mat_mult(x))
# With Center but No Scale:
# B = (A - U)
b_array = array - mu
result = b_array.dot(b_array.T.dot(x))/f
sarray = ScaledCenterArray(scale=False, center=True, factor=factor)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.sym_mat_mult(x))
# With Center and Scale:
# (A - U)'D'D(A - U)x
result = (array - mu).dot(std).dot(std).dot((array - mu).T.dot(x))/f
sarray = ScaledCenterArray(scale=True, center=True, factor=factor)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.sym_mat_mult(x))
def test_ScaledArray_different_x():
for N in range(2, 5):
for P in range(2, 5):
array = np.random.rand(N, P) + 1
std = np.diag(1/np.std(array, axis=0))
mu = np.mean(array, axis=0)
for K in range(3, 5):
x = np.random.rand(N, K)
for fit_x in [x, None]:
for dx in [-1, 0, 1]:
x2_K = x.shape[1] + dx
x2 = x[:, 0:x2_K]
# With No Scale or Center
# x = A'Ax
result = array.dot(array.T.dot(x2))
sarray = ScaledCenterArray(scale=False, center=False)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_equal(result, sarray.sym_mat_mult(x2))
# With Scale but No Center
# B = AD
b_array = array.dot(std)
result = b_array.dot(b_array.T.dot(x2))
sarray = ScaledCenterArray(scale=True, center=False)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.sym_mat_mult(x2))
# With Center but No Scale:
# B = (A - U)
b_array = array - mu
result = b_array.dot(b_array.T.dot(x2))
sarray = ScaledCenterArray(scale=False, center=True)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.sym_mat_mult(x2))
# With Center and Scale:
# (A - U)'D'D(A - U)x
result = (array - mu).dot(std).dot(std).dot((array - mu).T.dot(x2))
sarray = ScaledCenterArray(scale=True, center=True)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.sym_mat_mult(x2))
def test_ScaledArray_T_dot():
for N in range(2, 5):
for P in range(2, 5):
array = np.random.rand(N, P) + 1
std = np.diag(1/np.std(array, axis=0))
mu = np.mean(array, axis=0)
for K in range(1, 5):
for squeeze in [True, False]:
x = np.random.rand(N, K)
if squeeze:
x = np.squeeze(x)
for fit_x in [x, None]:
# With No Scale or Center
# x = A'Ax
result = array.T.dot(x)
sarray = ScaledCenterArray(scale=False, center=False)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.T.dot(x))
# With Scale but No Center
# B = AD
b_array = array.dot(std).T
result = b_array.dot(x)
sarray = ScaledCenterArray(scale=True, center=False)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.T.dot(x))
# With Center but No Scale:
# B = (A - U)
b_array = (array - mu).T
result = b_array.dot(x)
sarray = ScaledCenterArray(scale=False, center=True)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.T.dot(x))
# With Center and Scale:
# (A - U)'D'D(A - U)x
b_array = (array - mu).dot(std).T
result = b_array.dot(x)
sarray = ScaledCenterArray(scale=True, center=True)
sarray.fit(da.array(array), x=fit_x)
np.testing.assert_array_almost_equal(result, sarray.T.dot(x))