physbo.gp.core package

Submodules

• physbo.gp.core.learning module
  • Adam
    • Adam.get_one_update()
    • Adam.reset()
  • Batch
    • Batch.init_params_search()
    • Batch.one_run()
    • Batch.run()
  • Online
    • Online.disp_marlik()
    • Online.get_one_update()
    • Online.init_params_search()
    • Online.one_run()
    • Online.run()
  • adam()
  • batch()
  • online()

Module contents

class physbo.gp.core.Model(lik, mean, cov, inf='exact')

ベースクラス: object

パラメータ:

• lik

• mean

• cov

• inf

cat_params(lik_params, prior_params)

Concatinate the likelihood and prior parameters

パラメータ:

• lik_params (numpy.ndarray) -- Parameters for likelihood

• prior_params (numpy.ndarray) -- Parameters for prior

戻り値:

params -- parameters about likelihood and prior

戻り値の型:

numpy.ndarray

decomp_params(params=None)

decomposing the parameters to those of likelifood and priors

パラメータ:

params (numpy.ndarray) -- parameters

戻り値:

• lik_params (numpy.ndarray)

• prior_params (numpy.ndarray)

eval_marlik(params, X, t, N=None)

Evaluating marginal likelihood.

パラメータ:

• params (numpy.ndarray) -- Parameters.

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• N (int) -- Total number of subset data (if not specified, all dataset is used)

戻り値:

• marlik (float)

• Marginal likelihood.

export_blm(num_basis, comm=None)

Exporting the blm(Baysean linear model) predictor

パラメータ:

• num_basis (int) -- Total number of basis

• comm (MPI.Comm) -- MPI communicator

戻り値:

Bayesian linear model

戻り値の型:

physbo.blm.core.Model

fit(X, t, config, comm=None)

Fitting function (update parameters)

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• config (physbo.misc.set_config object)

• comm (MPI.Comm) -- MPI communicator

get_cand_params(X, t)

Getting candidate for parameters

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

戻り値:

params -- Parameters

戻り値の型:

numpy.ndarray

get_grad_marlik(params, X, t, N=None)

Evaluating gradiant of marginal likelihood.

パラメータ:

• params (numpy.ndarray) -- Parameters.

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• N (int) -- Total number of subset data (if not specified, all dataset is used)

戻り値:

grad_marlik -- Gradiant of marginal likelihood.

戻り値の型:

numpy.ndarray

get_params_bound()

Getting boundary of the parameters.

戻り値:

bound -- An array with the tuple (min_params, max_params).

戻り値の型:

list

get_permutation_importance(X, t, n_perm: int, comm=None, split_features_parallel=False)

Calculating permutation importance of model

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• n_perm (int) -- Number of permutations

• comm (MPI.Comm) -- MPI communicator

戻り値:

• numpy.ndarray -- importance_mean

• numpy.ndarray -- importance_std

get_post_fcov(X, Z, params=None, diag=True)

Calculating posterior covariance matrix of model (function)

パラメータ:

• X (numpy.ndarray) -- inputs

• Z (numpy.ndarray) -- feature maps

• params (numpy.ndarray) -- Parameters

• diag (bool) -- If true, only variances (diagonal elements) are returned.

戻り値:

Returned shape is (num_points) if diag=true, (num_points, num_points) if diag=false, where num_points is the number of points in X.

戻り値の型:

numpy.ndarray

get_post_fmean(X, Z, params=None)

Calculating posterior mean of model (function)

パラメータ:

• X (numpy.ndarray) -- inputs

• Z (numpy.ndarray) -- feature maps

• params (numpy.ndarray) -- Parameters

参考

physbo.gp.inf.exact.get_post_fmean

post_sampling(X, Z, params=None, N=1, alpha=1)

draws samples of mean value of model

パラメータ:

• X (numpy.ndarray) -- inputs

• Z (numpy.ndarray) -- feature maps

• N (int) -- number of samples (default: 1)

• alpha (float) -- noise for sampling source

戻り値の型:

numpy.ndarray

predict_sampling(X, Z, params=None, N=1)

パラメータ:

• X (numpy.ndarray) -- training datasets

• Z (numpy.ndarray) -- input for sampling objective values

• params (numpy.ndarray) -- Parameters

• N (int) -- number of samples (default: 1)

戻り値の型:

numpy.ndarray

prepare(X, t, params=None)

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• params (numpy.ndarray) -- Parameters.

print_params()

Printing parameters

set_params(params)

Setting parameters

パラメータ:

params (numpy.ndarray) -- Parameters.

sub_sampling(X, t, N)

Make subset for sampling

パラメータ:

• X (numpy.ndarray) -- Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- The negative energy of each search candidate (value of the objective function to be optimized).

• N (int) -- Total number of data in subset

戻り値:

• subX (numpy.ndarray)

• subt (numpy.ndarray)

class physbo.gp.core.Prior(mean, cov)

ベースクラス: object

prior of gaussian process

パラメータ:

• mean (physbo.gp.mean.Const) -- mean function

• cov (physbo.gp.cov.Gauss) -- covariance function

cat_params(mean_params, cov_params)

パラメータ:

• mean_params (numpy.ndarray) -- Mean values of parameters

• cov_params (numpy.ndarray) -- Covariance matrix of parameters

戻り値の型:

numpy.ndarray

decomp_params(params)

decomposing the parameters to those of mean values and covariance matrix for priors

パラメータ:

params (numpy.ndarray) -- parameters

戻り値:

• mean_params (numpy.ndarray)

• cov_params (numpy.ndarray)

get_cov(X, Z=None, params=None, diag=False)

Calculating the variance-covariance matrix of priors

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• Z (numpy.ndarray) -- N x d dimensional matrix. Each row of Z denotes the d-dimensional feature vector of tests.

• params (numpy.ndarray) -- Parameters.

• diag (bool) -- If true, only variances (diagonal elements) are returned.

戻り値:

Returned shape is (num_points) if diag=true, (num_points, num_points) if diag=false, where num_points is the number of points in X.

戻り値の型:

numpy.ndarray

get_grad_cov(X, params=None)

Calculating the covariance matrix priors

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• params (numpy.ndarray) -- Parameters.

戻り値の型:

numpy.ndarray

get_grad_mean(num_data, params=None)

Calculating the gradiant of mean values of priors

パラメータ:

• num_data (int) -- Total number of data

• params (numpy.ndarray) -- Parameters

戻り値の型:

numpy.ndarray

get_mean(num_data, params=None)

Calculating the mean value of priors

パラメータ:

• num_data (int) -- Total number of data

• params (numpy.ndarray) -- Parameters

戻り値の型:

float

sampling(X, N=1)

Sampling from GP prior

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• N (int)

戻り値の型:

float

set_cov_params(params)

Setting parameters for covariance matrix of priors

パラメータ:

params (numpy.ndarray) -- Parameters

set_mean_params(params)

Setting parameters for mean values of priors

パラメータ:

params (numpy.ndarray) -- Parameters

set_params(params)

Setting parameters

パラメータ:

params (numpy.ndarray) -- Parameters.

class physbo.gp.core.SFS(lik, mean, cov, inf='exact', config=None)

ベースクラス: Model

パラメータ:

• lik

• mean

• cov

• inf

fit(X, t)

Fitting function (update parameters)

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• config (physbo.misc.SetConfig object)

get_params(deep=True)

get_post_fmean(X, Z, params=None)

Calculating posterior mean of model (function)

パラメータ:

• X (numpy.ndarray) -- inputs

• Z (numpy.ndarray) -- feature maps

• params (numpy.ndarray) -- Parameters

参考

physbo.gp.inf.exact.get_post_fmean

predict(Z, params=None)

Calculating posterior mean of model (function)

パラメータ:

• X (numpy.ndarray) -- inputs

• Z (numpy.ndarray) -- feature maps

• params (numpy.ndarray) -- Parameters

参考

physbo.gp.inf.exact.get_post_fmean

prepare(X, t, params=None)

パラメータ:

• X (numpy.ndarray) -- N x d dimensional matrix. Each row of X denotes the d-dimensional feature vector of search candidate.

• t (numpy.ndarray) -- N dimensional array. The negative energy of each search candidate (value of the objective function to be optimized).

• params (numpy.ndarray) -- Parameters.