mgplvm.manifolds.torus module

class mgplvm.manifolds.torus.Torus(m, d)[source]

Bases: mgplvm.manifolds.base.Manifold

static distance(x, y, ell=None)[source]

Return type: Tensor

static expmap(x)[source]

move to [-pi, pi]

Return type: Tensor

static gmul(x, y)[source]

Return type: Tensor

inducing_points(n, n_z, z=None)[source]

static initialize(initialization, n_samples, m, d, Y)[source]: initializes latents - can add more exciting initializations as well

static inverse(x)[source]

Return type: Tensor

static log_q(log_base_prob, x, d, kmax)[source]: log of the variational distribution (~-H(Q))

property log_uniform: torch.Tensor

Return type: Tensor

static logmap(x)[source]

move to [-pi, pi]

Return type: Tensor

lprior(g)[source]

Return type: Tensor

property name

static parameterise(x)[source]

Return type: Tensor