Parallel Numerical Derivatives, Gradients, Jacobians, and Hessians of Arbitrary Accuracy Order

[Up] [Top]

Documentation for package ‘pnd’ version 0.0.9

Help Pages

checkCores Number of core checks and changes
checkDimensions Determine function dimensionality and vectorisation
dupRowInds Repeated indices of the first unique value
fdCoef Finite-difference coefficients for arbitrary grids
GenD Numerical derivative matrices with parallel capabilities
generateGrid Create a grid of points for a gradient / Jacobian
generateGrid2 Generate grid points for Hessians
Grad Gradient computation with parallel capabilities
gradstep Automatic step selection for numerical derivatives
Hessian Numerical Hessians
Jacobian Jacobian matrix computation with parallel capabilities s Computes the numerical Jacobian for vector-valued functions. Its columns are partial derivatives of the function with respect to the input elements. This function supports both two-sided (central, symmetric) and one-sided (forward or backward) derivatives. It can utilise parallel processing to accelerate computation of gradients for slow functions or to attain higher accuracy faster. Currently, only Mac and Linux are supported 'parallel::mclapply()'. Windows support with 'parallel::parLapply()' is under development.
plotTE Estimated total error plot as in Mathur (2012)
runParallel Run a function in parallel over a list (internal use only)
solveVandermonde Numerically stable non-confluent Vandermonde system solver
step.CR Curtis-Reid automatic step selection
step.DV Dumontet-Vignes automatic step selection
step.M Mathur's AutoDX-like automatic step selection
step.plugin Plug-in step selection
step.SW Stepleman-Winarsky automatic step selection
stepx Default step size at given points