rcalibration-class {RobustCalibration}

Robust Calibration class

Description

S4 class for Robust rcalibration with or without the specification of the discrepancy model.

Objects from the Class

Objects of this class are created and initialized with the function rcalibration that computes the calculations needed for setting up the calibration and prediction.

Slots

p_x:

Object of class integer. The dimension of the observed inputs.

p_theta:

Object of class integer. The calibration parameters.

num_obs:

Object of class integer. The number of experimental observations.

input:

Object of class matrix with dimension n x p_x. The design of experiments.

output:

Object of class vector with dimension n x 1. The vector of the experimental observations.

X:

Object of class matrix of with dimension n x q. The mean/trend discrepancy basis function.

have_trend:

Object of class bool to specify whether the mean/trend discrepancy is zero. "TRUE" means it has zero mean discrepancy and "FALSE"" means the mean discrepancy is not zero.

q:

Object of class integer. The number of basis functions of the mean/trend discrepancy.

R0:

Object of class list of matrices where the j-th matrix is an absolute difference matrix of the j-th input vector.

kernel_type:

A character to specify the type of kernel to use.

alpha:

Object of class vector. Each element is the parameter for the roughness for each input coordinate in the kernel.

theta_range:

A matrix for the range of the calibration parameters.

lambda_z:

Object of class vector about how close the math model to the reality in squared distance when the S-GaSP model is used for modeling the discrepancy.

S:

Object of class integer about how many posterior samples to run.

S_0:

Object of class integer about the number of burn-in samples.

prior_par:

Object of class vector about prior parameters.

output_weights:

Object of class vector about the weights of the experimental data.

sd_proposal:

Object of class vector about the standard deviation of the proposal distribution.

discrepancy_type:

Object of class character about the discrepancy. If it is 'no-discrepancy', it means no discrepancy function. If it is 'GaSP', it means the GaSP model for the discrepancy function. If it is 'S-GaSP', it means the S-GaSP model for the discrepancy function.

simul_type:

Object of class integer about the math model/simulator. If the simul_type is 0, it means we use RobustGaSP R package to build an emulator for emulation. If the simul_type is 1, it means the function of the math model is given by the user. When simul_type is 2 or 3, the mathematical model is the geophyiscal model for Kilauea Volcano. If the simul_type is 2, it means it is for the ascending mode InSAR data; if the simul_type is 3, it means it is for the descending mode InSAR data.

emulator_rgasp:

An S4 class of rgasp from the RobustGaSP package.

emulator_ppgasp:

An S4 class of ppgasp from the RobustGaSP package.

post_sample:

Object of class matrix for the posterior samples after burn-in.

post_value:

Object of class vector for the posterior values after burn-in.

accept_S:

Object of class vector for the number of proposed samples of the calibation parameters are accepted in MCMC. The first value is the number of proposed calibration parameters are accepted in MCMC. The second value is the number of proposed range and nugget parameters are accepted.

count_boundary:

Object of class vector for the number of proposed samples of the calibation parameters are outside the range and they are rejected directly.

have_replicates:

Object of class bool for having repeated experiments (replicates) or not.

num_replicates:

Object of class vector for the number of replicates at each observable input.

thinning:

Object of class integer for the ratio between the number of posterior samples and the number of samples to be recorded.

S_2_f:

Object of class numeric for the variance of the field observations.

num_obs_all:

Object of class integer for the total number of field observations.

method:

Object of class character for posterior sampling or maximum likelihood estimation.

initial_values:

Object of class matrix for initial starts of kernel parameters in maximum likelihood estimation.

param_est:

Object of class vector for estimated range and nugget parameter in parameter estimation.

opt_value:

Object of class numeric for optimized likelihood or loss function.

emulator_type:

Object of class character for the type of emulator. 'rgasp' means scalar-valued emulator and 'ppgasp' means vectorized emulator.

loc_index_emulator:

Object of class vector for location index to output in the ppgasp emulator for computer models with vectorized output.

Methods

show

Prints the main slots of the object.

predict

See predict.

predict_separable_2dim

See predict_separable_2dim.

Author(s)

Mengyang Gu [aut, cre]

Maintainer: Mengyang Gu <[e-mail]>

References

A. O'Hagan and M. C. Kennedy (2001), Bayesian calibration of computer models, Journal of the Royal Statistical Society: Series B (Statistical Methodology, 63, 425-464.

M. Gu (2016), Robust Uncertainty Quantification and Scalable Computation for Computer Models with Massive Output, Ph.D. thesis., Duke University.

M. Gu and L. Wang (2017) Scaled Gaussian Stochastic Process for Computer Model Calibration and Prediction. arXiv preprint arXiv:1707.08215.

M. Gu (2018) Jointly Robust Prior for Gaussian Stochastic Process in Emulation, Calibration and Variable Selection . arXiv preprint arXiv:1804.09329.

See Also

rcalibration for more details about how to create a rcalibration object.