fit_linear_model_of_coregionalization: Fit Linear Model of Coregionalization Ported from Emergy et...

In ScottHMcKean/duvernay_geomech_model: A Geostatistical Model of Geomechanical Properties in the Duvernay Play

Fit Linear Model of Coregionalization Ported from Emergy et al. 2009

Description

Fit Linear Model of Coregionalization Ported from Emergy et al. 2009

Usage

fit_linear_model_of_coregionalization(
  azimuth,
  dip,
  nlag,
  tail,
  head,
  gam,
  model,
  vartype,
  maxiterations = 5000,
  max_time_s = 100,
  p_acc_0 = 0.9,
  p_acc_2k = 0.1,
  gen_vec_corr = 0.9,
  eps_thresh = 1e-12,
  weighting_method = 1
)

Arguments

azimuth	azimuth for sample variogram/covariance (nvariogram * 1 vector), degrees
dip	dips for sample variogram/covariance (nvariog * 1 vector), degrees
nlag	number of lags for each variogram/correlogram (nvariog * 1 vector)
tail	tail variables (nvariog * 1 vector), an integer vector
head	head variables (nvariog * 1 vector), an integer vector
gam	dataframe of simple and cross variograms/covariances. The shape of the dataframe should be sum(nlag) rows x 3 columns
model	the variogram/covariance model matrix with dimensions of the number of structures (n_structures) * 8
vartype	the script handles traditional variograms (1) or centered covariances (2)
maxiterations	maximum number of consecutive iterations with no accepted transitions
max_time_s	maximum processing time in seconds
p_acc_0	probability of accepting a non-favorable transition at iteration 0
p_acc_2k	probability of accepting a non-favorable transition at iteration 2000
gen_vec_corr	correlation between generated Gaussian vectors
eps_thresh	absolute comparison threshold for machine accuracy)
weighting_method	0:constant weight, 1: ~npairs, 2:~/distance 3: ~npairs/distance

Value

sills of nested structures (n_structures * n_fields^2 matrix) and weighted sum of squares for optimal fit

ScottHMcKean/duvernay_geomech_model documentation built on Sept. 12, 2022, 10:08 a.m.