AffineTransform

class pyapprox.variables.AffineTransform(variable, enforce_bounds=False)

Bases: object

Apply an affine transformation to a pyapprox.variables.IndependentMarginalsVariable

Methods Summary

get_ranges()
map_derivatives_from_canonical_space(derivatives)	Parameters:
map_derivatives_to_canonical_space(...)	derivatives : np.ndarray (nvars*nsamples, nqoi)
map_from_canonical(canonical_samples)
map_from_canonical_1d(canonical_samples, ii)
map_to_canonical(user_samples)
map_to_canonical_1d(samples, ii)
num_vars()
samples_of_bounded_variables_inside_domain(samples)
set_identity_maps(identity_map_indices)	Set the dimensions we do not want to map to and from canonical space

Methods Documentation

get_ranges()

map_derivatives_from_canonical_space(derivatives)

Parameters:

derivativesnp.ndarray (nvars*nsamples, nqoi)

Derivatives of each qoi. The ith column consists of the derivatives [d/dx_1 f(x^{(1)}), …, f(x^{(M)}), d/dx_2 f(x^{(1)}), …, f(x^{(M)}) …, d/dx_D f(x^{(1)}), …, f(x^{(M)})] where M is the number of samples and D=nvars

Derivatives can also be (nvars, nsamples) - transpose of Jacobian - Here each sample is considered a different QoI

map_derivatives_to_canonical_space(canonical_derivatives)

derivativesnp.ndarray (nvars*nsamples, nqoi)

Derivatives of each qoi. The ith column consists of the derivatives [d/dx_1 f(x^{(1)}), …, f(x^{(M)}),

d/dx_2 f(x^{(1)}), …, f(x^{(M)}) …, d/dx_D f(x^{(1)}), …, f(x^{(M)})]

where M is the number of samples and D=nvars

Derivatives can also be (nvars, nsamples) - transpose of Jacobian - Here each sample is considered a different QoI

map_from_canonical(canonical_samples)

map_from_canonical_1d(canonical_samples, ii)

map_to_canonical(user_samples)

map_to_canonical_1d(samples, ii)

num_vars()

samples_of_bounded_variables_inside_domain(samples)

set_identity_maps(identity_map_indices)

Set the dimensions we do not want to map to and from canonical space

Parameters:

identity_map_indicesiterable

The dimensions we do not want to map to and from canonical space