GaussianMeasure

class probnum.quad.GaussianMeasure(mean, cov, input_dim=None)

Bases: IntegrationMeasure

Gaussian measure on Euclidean space with given mean and covariance.

If mean and cov are scalars but input_dim is larger than one, mean and cov are extended to a constant vector and diagonal matrix, respectively, of appropriate dimensions.

Parameters

  • mean (Union[float, np.floating, np.ndarray]) – shape=(input_dim,) – Mean of the Gaussian measure.

  • cov (Union[float, np.floating, np.ndarray]) – shape=(input_dim, input_dim) – Covariance matrix of the Gaussian measure.

  • input_dim (Optional[IntLike]) – Dimension of the integration domain.

Return type

None

Methods Summary

__call__(points)

Evaluate the density function of the integration measure.

sample(n_sample[, rng])

Sample n_sample points from the integration measure.

Methods Documentation

__call__(points)

Evaluate the density function of the integration measure.

Parameters

points (Union[FloatLike, np.ndarray]) – shape=(n_points, input_dim) – Input locations.

Returns

shape=(n_points,) – Density evaluated at given locations.

Return type

density_evals

sample(n_sample, rng=Generator(PCG64) at 0x7F73483C9040)

Sample n_sample points from the integration measure.

Parameters

  • n_sample (IntLike) – Number of points to be sampled

  • rng (Optional[Generator]) – Random number generator. Optional. Default is np.random.default_rng().

Returns

shape=(n_sample,input_dim) – Sampled points

Return type

points