Description Details Author(s) References See Also Examples
Precise numerical evaluation of arbitrary order derivatives for functions of the form $f:R^m \to R^n$ through both forward and reverse mode Automatic Differentiation
The main functions are:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | radx Create objects of class radx. These store the truncated
Taylor series expansions which are propagated through
expressions to compute derivatives.
getDeriv Obtain the derivative from the taylor series coefficients
of a radx class expression.
radxeval Driver for computing arbitrary order derivatives of functions.
Yields a matrix with derivatives of each output function along
one column ordered by the multi-index corresponding to the
derivative they represent.
convert2pos Convert a given set of multi-indices (a matrix with one
multi-index per row) to a vector of positional indices indicating
the location of the corresponding derivative in the output of radxeval.
Example: The mixed derivative (d^3 / dx^2 dy^1 dz^0)(f) has the
multi-index (2, 1, 0) whose positional index is 2. Thus, this
mixed derivative can be found along the second row of the matrix
in the output of radxeval.
|
Maintainer: Chidambaram Annamalai <quantumelixir@gmail.com>
Griewank, A. and Walther, A. (2008) Evaluating derivatives: principles and techniques of algorithmic differentiation. Society for Industrial and Applied Mathematics (SIAM)
Griewank, A. and Utke, J. and Walther, A., (2000) Evaluating higher derivative tensors by forward propagation of univariate Taylor series. Journal of Mathematics of Computation, American Mathematical Society.
1 2 3 4 5 6 7 8 9 | ## Not run:
## run demos
demo("univariate")
demo("hessian")
## open the directory with documents
browseURL(paste(system.file(package="radx"), "/doc", sep=""))
## End(Not run)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.