Package: pnd 0.0.9

Andreï Victorovitch Kostyrka

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

Numerical derivatives through finite-difference approximations can be calculated using the 'pnd' package with parallel capabilities and optimal step-size selection to improve accuracy. These functions facilitate efficient computation of derivatives, gradients, Jacobians, and Hessians, allowing for more evaluations to reduce the mathematical and machine errors. Designed for compatibility with the 'numDeriv' package, which has not received updates in several years, it introduces advanced features such as computing derivatives of arbitrary order, improving the accuracy of Hessian approximations by avoiding repeated differencing, and parallelising slow functions on Windows, Mac, and Linux.

Authors:Andreï Victorovitch Kostyrka [aut, cre]

pnd_0.0.9.tar.gz
pnd_0.0.9.zip(r-4.5)pnd_0.0.9.zip(r-4.4)pnd_0.0.9.zip(r-4.3)
pnd_0.0.9.tgz(r-4.5-any)pnd_0.0.9.tgz(r-4.4-any)pnd_0.0.9.tgz(r-4.3-any)
pnd_0.0.9.tar.gz(r-4.5-noble)pnd_0.0.9.tar.gz(r-4.4-noble)
pnd_0.0.9.tgz(r-4.4-emscripten)pnd_0.0.9.tgz(r-4.3-emscripten)
pnd.pdf |pnd.html
pnd/json (API)
NEWS

# Install 'pnd' in R:
install.packages('pnd', repos = c('https://fifis.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/fifis/pnd/issues

On CRAN:

Conda:

5.22 score 1 stars 5 scripts 240 downloads 20 exports 2 dependencies

Last updated 3 days agofrom:6d49d67cec. Checks:9 OK. Indexed: yes.

TargetResultLatest binary
Doc / VignettesOKMar 11 2025
R-4.5-winOKMar 11 2025
R-4.5-macOKMar 11 2025
R-4.5-linuxOKMar 11 2025
R-4.4-winOKMar 11 2025
R-4.4-macOKMar 11 2025
R-4.4-linuxOKMar 11 2025
R-4.3-winOKMar 11 2025
R-4.3-macOKMar 11 2025

Exports:checkCorescheckDimensionsdupRowIndsfdCoefGenDgenerateGridgenerateGrid2GradgradstepHessianJacobianplotTErunParallelsolveVandermondestep.CRstep.DVstep.Mstep.pluginstep.SWstepx

Dependencies:rbibutilsRdpack

Compatilibility of pnd with the syntax of numDeriv

Rendered fromcompatibility-with-numDeriv.Rmdusingknitr::rmarkdownon Mar 11 2025.

Last update: 2025-03-11
Started: 2024-06-26

Fast and accurate parallel numerical derivatives in R

Rendered fromfast-and-accurate.Rmdusingknitr::rmarkdownon Mar 11 2025.

Last update: 2025-01-24
Started: 2024-06-13

Step-size-selection algorithm benchmark

Rendered fromstep-size-selection.Rmdusingknitr::rmarkdownon Mar 11 2025.

Last update: 2025-01-24
Started: 2024-06-13

Readme and manuals

Help Manual

Help pageTopics
Number of core checks and changescheckCores
Determine function dimensionality and vectorisationcheckDimensions
Repeated indices of the first unique valuedupRowInds
Finite-difference coefficients for arbitrary gridsfdCoef
Numerical derivative matrices with parallel capabilitiesGenD
Create a grid of points for a gradient / JacobiangenerateGrid
Generate grid points for HessiansgenerateGrid2
Gradient computation with parallel capabilitiesGrad
Automatic step selection for numerical derivativesgradstep
Numerical HessiansHessian
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.Jacobian
Estimated total error plot as in Mathur (2012)plotTE
Run a function in parallel over a list (internal use only)runParallel
Numerically stable non-confluent Vandermonde system solversolveVandermonde
Curtis-Reid automatic step selectionstep.CR
Dumontet-Vignes automatic step selectionstep.DV
Mathur's AutoDX-like automatic step selectionstep.M
Plug-in step selectionstep.plugin
Stepleman-Winarsky automatic step selectionstep.SW
Default step size at given pointsstepx