def _impl_test_add_diag(self, dtype):
x = np.asarray(np.random.rand(4, 4), dtype)
d = np.asarray(np.random.rand(1, 4), dtype).reshape(-1)
x_gpu = gpuarray.to_gpu(x)
d_gpu = gpuarray.to_gpu(d)
res_cpu = x + np.diag(d)
res_gpu = linalg.add_diag(d_gpu, x_gpu, overwrite=False)
assert np.allclose(res_cpu, res_gpu.get(), atol=1e-5)
assert res_gpu is not x_gpu
res_gpu = linalg.add_diag(d_gpu, x_gpu, overwrite=True)
assert np.allclose(res_cpu, res_gpu.get(), atol=1e-5)
assert res_gpu is x_gpu
def impl_test_add_diag(self, dtype):
x = np.asarray(np.random.rand(4, 4), dtype)
d = np.asarray(np.random.rand(1, 4), dtype).reshape(-1)
x_gpu = gpuarray.to_gpu(x)
d_gpu = gpuarray.to_gpu(d)
res_cpu = x + np.diag(d)
res_gpu = linalg.add_diag(d_gpu, x_gpu, overwrite=False)
assert np.allclose(res_cpu, res_gpu.get(), atol=1e-5)
assert res_gpu is not x_gpu
res_gpu = linalg.add_diag(d_gpu, x_gpu, overwrite=True)
assert np.allclose(res_cpu, res_gpu.get(), atol=1e-5)
assert res_gpu is x_gpu
def calculate_H_gpu(X, W, P):
WPW = la.add_diag(P, la.dot(W, W, "t", "n"))
tmp = la.dot(W, la.inv(WPW, overwrite=True))
H = la.dot(X, tmp, "n", "t")
H = gpu.maximum(H, 0)
H = to_unit_variance(H)
return H, tmp
def calculate_H_gpu(X, W, P):
WPW = la.add_diag(P, la.dot(W, W, "t", "n"))
tmp = la.dot(W, la.inv(WPW, overwrite=True))
H = la.dot(X, tmp, "n", "t")
H = gpu.maximum(H, 0)
H = to_unit_variance(H)
return H, tmp
