SG++-Doxygen-Documentation
python.painlesscg Namespace Reference

Functions

def ApplyA (B, C, alpha, result, x, l)
 
def BiCGStab (b, alpha, imax, epsilon, ApplyMatrix, verbose=True)
 
def cg (y, alpha, grid, x, imax, epsilon, l, verbose=True)
 
def cg_new (b, alpha, imax, epsilon, ApplyMatrix, reuse=False, verbose=True, max_threshold=None)
 Conjugated Gradient method for sparse grids, solving A.alpha=b. More...
 
def sd (y, alpha, grid, x, imax, epsilon, l)
 

Function Documentation

def python.painlesscg.ApplyA (   B,
  C,
  alpha,
  result,
  x,
  l 
)
def python.painlesscg.BiCGStab (   b,
  alpha,
  imax,
  epsilon,
  ApplyMatrix,
  verbose = True 
)
def python.painlesscg.cg (   y,
  alpha,
  grid,
  x,
  imax,
  epsilon,
  l,
  verbose = True 
)
def python.painlesscg.cg_new (   b,
  alpha,
  imax,
  epsilon,
  ApplyMatrix,
  reuse = False,
  verbose = True,
  max_threshold = None 
)

Conjugated Gradient method for sparse grids, solving A.alpha=b.

The resulting vector is stored in alpha.

Parameters
bRHS of equation
alphavector of unknowns
imaxmax. number of iterations (abort, if reached)
epsilonaccuracy requirements (reduce initial norm of residuum |delta_0| below epsilon*|delta_0|)
ApplyMatrixprocedure that applies A to a vector
reusestarting vector is 0 by default. If true, use current values in alpha
verboseverbose output (default False)
max_thresholdmaximal threshold
Returns
tuple (number of iterations, final norm of residuum)

Referenced by python.classifier.doTest(), python.classifier.performFold(), python.classifier.performFoldNew(), python.classifier.performFoldRegression(), and python.classifier.run().

def python.painlesscg.sd (   y,
  alpha,
  grid,
  x,
  imax,
  epsilon,
  l 
)