C++ Classes List

The following is a list of C++ classes that I have written and am distributing by request to those that want them. This list is changing frequently depending upon demand and my free time.

Derivs

Calculates the first derivative of a single parameter function by using Richardson Extrapolation. This class is a C++ translation of an Ada package in: Ochs, T. 1991; Building Blocks: Now you see it, now you don't, Computer Language, V.8 No. 5, May, pp. 97 -106

Fredholm2nd, Volterra2nd (shell archive, or GNU Compressed tar)

Classes for solving linear Fredholm and Volterra equations of the second kind.

GTri ( Unix shell archive , or GNU Compressed tar)

Class for solving Tridiagonal systems of linear equations. Uses the Thomas algorithm. Can use Dirichlet, Neumann, or Mixed boundary conditions.

Kalman ( Unix compressed format tar, or GNU Compressed tar)

A general purpose Kalman filter class for optimal estimation of a dynamical system. Updated 5 Sept 1997 with portability enhancements. Updated 27 Jan 2002 to simplify the model API.

KernelND, KernelNDAdaptive

Classes for making kernel estimates of a PDF in N dimensions. There is both a fixed and an adaptive form.

Model, Map, Ode, Runge

A family of classes that allows the treatement of iterative maps and ordinary differential equations in a unified manner. The ordinary differential equations are solved using the class Runge which is a C++ adaptation of the code given in: W.H. Press, B.P. Flannery, S.A. Teukolsky and W.T. Vetterling, Numerical Recipes: The Art of Scientific Computing, Cambridge Univ. Press, New York, 1986 These classes were the example classes described in my recent Computer Language article: The Evolution of a C++ programmer, Computer Language, V. 9, No. 8, August 1992

MonteIntegrator ( shar, tar) , Quasi ( shar, tar)

A Monte Carlo function integrator and its supporting random number generator. The random number generator is a quasi-random (as opposed to a pseudo-random) number generator, this gives the integrator a 1/N convergence rate (as opposed to 1/Sqrt(N) ). Can integrate functions up to 52 dimensions (this limit is due to the quasi random number generator).

MCInv ( shar, tar)

A library of routines that use Markov Chains for the estimation of: A single element of the inverse of a matrix. A given row of the inverse of a matrix. The solution of x = A x + f ( Matrix A, vectors x and f) for a given row. Based upon ACM Algorithm #166

Seawater, SeaParse

A pair of classes for handling hydrographic data.

Shared

A Unix Shared Memory object base class. When this class is inherited by other classes then the derived class becomes a shared memory object that can be accessed by multiple processes. These objects can be persistant.

SdeInt ( Unix compressed tar file,123K or other archive formats)

A Runge-Kutta like class for integrating systems of Stochastic Differential Equations. (includes the R250 random number generator).

SimAnneal

A general purpose Simulated Annealing class, for estimating optimum parameters of complicated functions or systems. (includes the R250 random number generator).

  
  Also of interest 
 

Index of free source code for numerical computation in C or C++ (an old local copy, let us know if you know where the official current one now is)

C++ FAQ

C++ Virtual Library

Taygeta's home page