reinitialize pc_model for splits

sandialabs · Jun 8, 2023 · 9ca2fc6 · 9ca2fc6
1 parent 8b61940
commit 9ca2fc6
Show file tree

Hide file tree

Showing 2 changed files with 59 additions and 58 deletions.
diff --git a/PyUQTk/PyPCE/pce_tools.py b/PyUQTk/PyPCE/pce_tools.py
@@ -371,7 +371,7 @@ def UQTkRegression(pc_model,f_evaluations, samplepts):
  # Return numpy array of PC coefficients
  return c_k
 ################################################################################
-def UQTkBCS(pc_model, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
+def UQTkBCS(pc_begin, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
  mindex_growth='nonconservative', regparams=None, sigma2=1e-8, trval_frac=None,\
  npccut=None, pcf_thr=None, verbose=0, eta_plot=False):
  """
@@ -381,7 +381,7 @@ def UQTkBCS(pc_model, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
  ToDo: add documentation in UQTk manual on what BCS is and the basis growth schemes
 
  Input:
- pc_model: PC object with information about the starting basis
+ pc_begin: PC object with information about the starting basis
  xdata: N-dimensional NumPy array with sample points [#samples,
  #dimensions]
  ydata: 1D numpy array (vector) with function, evaluated at the
@@ -428,7 +428,7 @@ def UQTkBCS(pc_model, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
  eta_opt = eta
  elif (type(eta)==np.ndarray or type(eta)==list):
  # the eta with the lowest RMSE is selected from etas
- eta_opt = UQTkOptimizeEta(pc_model, ydata, xdata, eta, niter, eta_folds, mindex_growth, verbose, eta_plot)
+ eta_opt = UQTkOptimizeEta(pc_begin, ydata, xdata, eta, niter, eta_folds, mindex_growth, verbose, eta_plot)
  if verbose:
  print("Optimal eta is", eta_opt)
  else:
@@ -448,12 +448,12 @@ def UQTkBCS(pc_model, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
  if regparams is None:
  regparams = np.array([])
  elif type(regparams)==int or type(regparams)==float:
- regparams = regparams*np.ones((pc_model.GetNumberPCTerms(),))
+ regparams = regparams*np.ones((pc_begin.GetNumberPCTerms(),))
 
  if mindex_growth == None:
- full_basis_size = pc_model.GetNumberPCTerms()
+ full_basis_size = pc_begin.GetNumberPCTerms()
  else:
- full_basis_size = uqtkpce.PCSet("NISPnoq", pc_model.GetOrder() + niter -1, pc_model.GetNDim(), pc_model.GetPCType(), pc_model.GetAlpha(), pc_model.GetBeta()).GetNumberPCTerms()
+ full_basis_size = uqtkpce.PCSet("NISPnoq", pc_begin.GetOrder() + niter -1, pc_begin.GetNDim(), pc_begin.GetPCType(), pc_begin.GetAlpha(), pc_begin.GetBeta()).GetNumberPCTerms()
 
  # loop through iterations with different splits of the data
  for i in range(ntry):
@@ -468,16 +468,16 @@ def UQTkBCS(pc_model, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
  # Iterations of multiindex growth
  for j in range(niter):
  # Retrieve multiindex
- mi_uqtk = uqtkarray.intArray2D(pc_model.GetNumberPCTerms(), nsam)
- pc_model.GetMultiIndex(mi_uqtk)
+ mi_uqtk = uqtkarray.intArray2D(pc_begin.GetNumberPCTerms(), nsam)
+ pc_begin.GetMultiIndex(mi_uqtk)
  mindex=uqtkarray.uqtk2numpy(mi_uqtk)
  if verbose>0:
  print("==== BCS with multiindex of size %d ====" % (mindex.shape[0],))
  if verbose>1:
  print(mindex)
 
  # One run of BCS to obtain an array of coefficients and a new multiindex
- c_k, used_mi_np = UQTkEvalBCS(pc_model, y_split, x_split, sigma2, eta_opt, regparams, verbose)
+ c_k, used_mi_np = UQTkEvalBCS(pc_begin, y_split, x_split, sigma2, eta_opt, regparams, verbose)
 
  # Custom 'cuts' by number of PC terms or by value of PC coefficients
  npcall = c_k.shape[0] # number of PC terms
@@ -524,8 +524,8 @@ def UQTkBCS(pc_model, xdata, ydata, niter, eta, ntry=1, eta_folds=5,\
  mindex_uq.assign(i2,j2, mindex[i2][j2])
 
  # create a pc object with the new multiindex
- pc_model=uqtkpce.PCSet("NISPnoq", mindex_uq, pc_model.GetPCType(),\
- pc_model.GetAlpha(), pc_model.GetBeta())
+ pc_model=uqtkpce.PCSet("NISPnoq", mindex_uq, pc_begin.GetPCType(),\
+ pc_begin.GetAlpha(), pc_begin.GetBeta())
 
  # Save for this trial
  mi_selected.append(mindex)