physbo.blm package

Subpackages

Module contents

Bayesian Linear Model

class physbo.blm.Model(lik, prior, options={})[ソース]

ベースクラス: object

Baysean Linear Model

prior

prior distribution of weights

Type:

physbo.blm.prior.Gauss

lik

kernel

Type:

physbo.blm.lik.Gauss

nbasis

number of features in random feature map

Type:

int

stats

auxially parameters for sampling

Type:

Tuple

method

sampling method

Type:

str

get_permutation_importance(X, t, n_perm: int, comm=None, split_features_parallel=False)[ソース]

Calculating permutation importance of model

パラメータ:

  • X (numpy.ndarray) -- inputs

  • t (numpy.ndarray) -- target (label)

  • 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, Psi=None, diag=True)[ソース]

calculates posterior covariance of model

パラメータ:

  • X (numpy.ndarray) -- inputs

  • Psi (numpy.ndarray) -- feature maps (default: blm.lik.linear.basis.get_basis(X))

  • diag (bool) -- if True, returns only variances as a diagonal matrix (default: True)

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

参考

physbo.blm.inf.exact.get_post_fcov

get_post_fmean(X, Psi=None, w=None)[ソース]

calculates posterior mean of model (function)

パラメータ:

  • X (numpy.ndarray) -- inputs

  • Psi (numpy.ndarray) -- feature maps

  • w (numpy.ndarray) -- weight

参考

physbo.blm.inf.exact.get_post_fmean

get_post_params_mean()[ソース]

calculates posterior mean of weights

戻り値の型:

numpy.ndarray

参考

physbo.blm.inf.exact.get_post_params_mean

post_sampling(Xtest, Psi=None, N=1, alpha=1.0)[ソース]

draws samples of mean value of model

パラメータ:

  • Xtest (numpy.ndarray) -- inputs

  • Psi (numpy.ndarray) -- feature maps (default: blm.lik.get_basis(Xtest))

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

  • alpha (float) -- noise for sampling source

戻り値の型:

numpy.ndarray

predict_sampling(Xtest, Psi=None, N=1)[ソース]

draws samples from model

パラメータ:

  • Xtest (numpy.ndarray) -- inputs

  • Psi (numpy.ndarray) -- feature map (default: blm.lik.get_basis(Xtest))

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

戻り値の型:

numpy.ndarray

prepare(X, t, Psi=None)[ソース]

initializes model by using the first training dataset

パラメータ:

  • X (numpy.ndarray) -- inputs

  • t (numpy.ndarray) -- target (label)

  • Psi (numpy.ndarray) -- feature maps

参考

physbo.blm.inf.exact.prepare

sampling(w_mu=None, N=1, alpha=1.0)[ソース]

draws samples of weights

パラメータ:

  • blm (physbo.blm.core.Model) -- model

  • w_mu (numpy.ndarray) -- mean of weight

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

  • alpha (float) -- noise for sampling source (default: 1.0)

戻り値:

samples of weights

戻り値の型:

numpy.ndarray

参考

physbo.blm.inf.exact.sampling

update_stats(x, t, psi=None)[ソース]

updates model by using another training data

パラメータ:

  • x (numpy.ndarray) -- input

  • t (float) -- target (label)

  • psi (numpy.ndarray) -- feature map

参考

physbo.blm.inf.exact.update_stats

class physbo.blm.Predictor(config, model=None)[ソース]

ベースクラス: BasePredictor

Predictor using Baysean linear model

blm
Type:

physbo.blm.core.Model

config

configuration

Type:

physbo.misc.SetConfig

パラメータ:

参考

physbo.predictor.BasePredictor

delete_stats()[ソース]

resets model

fit(training, num_basis=None, comm=None)[ソース]

fit model to training dataset

パラメータ:

  • training (physbo.variable) -- dataset for training

  • num_basis (int) -- the number of basis (default: self.config.predict.num_basis)

  • comm (MPI.Comm) -- MPI communicator

get_basis(X)[ソース]

calculates feature maps Psi(X)

パラメータ:

X (numpy.ndarray) -- inputs

戻り値:

Psi -- feature maps

戻り値の型:

numpy.ndarray

get_post_fcov(training, test, diag=True)[ソース]

calculates posterior variance-covariance matrix of model

パラメータ:

  • training (physbo.variable) -- training dataset. If already trained, the model does not use this.

  • test (physbo.variable) -- inputs

  • 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 test.

戻り値の型:

numpy.ndarray

get_post_fmean(training, test)[ソース]

calculates posterior mean value of model

パラメータ:

  • training (physbo.variable) -- training dataset. If already trained, the model does not use this.

  • test (physbo.variable) -- inputs

戻り値:

Returned shape is (num_points), where num_points is the number of points in test.

戻り値の型:

numpy.ndarray

get_post_params(training, test)[ソース]

calculates posterior weights

パラメータ:

  • training (physbo.variable) -- training dataset. If already trained, the model does not use this.

  • test (physbo.variable) -- inputs (not used)

戻り値の型:

numpy.ndarray

get_post_samples(training, test, N=1, alpha=1.0)[ソース]

draws samples of mean values of model

パラメータ:

  • training (physbo.variable) -- training dataset. If already trained, the model does not use this.

  • test (physbo.variable) -- inputs

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

  • alpha (float) -- noise for sampling source (default: 1.0)

戻り値の型:

numpy.ndarray

get_predict_samples(training, test, N=1)[ソース]

draws samples of values of model

パラメータ:

  • training (physbo.variable) -- training dataset. If already trained, the model does not use this.

  • test (physbo.variable) -- inputs

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

  • alpha (float) -- noise for sampling source (default: 1.0)

戻り値の型:

numpy.ndarray (N x len(test))

prepare(training)[ソース]

initializes model by using training data set

パラメータ:

training (physbo.variable) -- dataset for training

update(training, test)[ソース]

updates the model.

If not yet initialized (prepared), the model will be prepared by training. Otherwise, the model will be updated by test.

パラメータ:

  • training (physbo.variable) -- training dataset for initialization (preparation). If already prepared, the model ignore this.

  • test (physbo.variable) -- training data for update. If not prepared, the model ignore this.