Skip to content

Fast gradient evaluation in C++ based on Expression Templates.

License

Notifications You must be signed in to change notification settings

stillwater-sc/CoDiPack

 
 

Repository files navigation

CoDiPack

CoDiPack (Code Differentiation Package) is a tool for gradient evaluation in computer programs. It supports the features:

  • Forward mode of Algorithmic Differentiation (AD)
  • Reverse mode of Algorithmic Differentiation (AD)
  • Different tape implementations
  • An AdjointMPI interface
  • External functions
  • Higher order derivatives

The design principle for CoDiPack is that it is easy to use. However, it also gives experienced AD developers full access to all the data structures.

The Scientific Computing Group at the TU Kaiserslautern develops CoDiPack and will enhance and extend CoDiPack in the future. There is a newsletter available at codi-info@uni-kl.de and if you want to contact us please write a mail to codi@scicomp.uni-kl.de.

Build Status DOI

Usage

CoDiPack is a header only library. The only file the user needs to include is codi.hpp. The only other requirement is a C++11 compliant compiler where one usually needs to specify '-std=c++11' in compiler arguments. CoDiPack is tested with gcc, clang, and the Intel compiler.

The file codi.hpp defines several datatypes. The most important ones are:

  • Implementations of the forward mode of AD:
    • codi::RealForward
  • Implementation of the reverse mode of AD:
    • codi::RealReverse (most common type, works everywhere, C-compatible)
    • codi::RealReverseIndex (reduced tape size w.r.t. codi::RealReverse, no C-like memory operations (e.g. memcpy))
    • codi::RealReversePrimal (reduced tape size w.r.t. codi::RealReverseIndex, works everywhere, C-compatible, increased type complexity)
    • codi::RealReversePrimalIndex (reduced tape size w.r.t. codi::RealReversePrimal, no C-like memory operations (e.g. memcpy), increased type complexity)

We recommend to use the codi::RealReverse type when AD is first introduced to an application. After that there should be no difficulties in replacing the codi::RealReverse type with other types.

For further details please visit our CoDiPack web page.

Miscellaneous information

Debugging with gdb

The ActiveReal type contains the tape as a static member. GDB prints the information of these members in its default settings, which makes the output quite verbose. We recommend to disable the output of the static class members. This can be done with

set print static-members off

Intel compiler options

Because CoDiPack relies on inlining of the compiler the performance can drop if it is not done or ignored. Therefore we recomend to force inlining of CoDiPack with the option

-DCODI_UseForcedInlines 

Hello World Example

A very small and simple example for the usage of the RealForward type is the following code:

    #include <codi.hpp>
    #include <iostream>

    int main(int nargs, char** args) {
      codi::RealForward x = 4.0;
      x.setGradient(1.0);

      codi::RealForward y = x * x;

      std::cout << "f(4.0) = " << y << std::endl;
      std::cout << "df/dx(4.0) = " << y.getGradient() << std::endl;

      return 0;
    }

It is compiled with

  g++  -I<path to codi>/include -std=c++11 -g -o forward forward.cpp

for the gcc compiler or with

  icpc  -I<path to codi>/include -std=c++11 -g -o forward forward.cpp

for the Intel compiler.

Please visit the tutorial page for further information.

Citation

If you use CoDiPack in one of your applications and write a paper it would be nice if you could cite the paper High-Performance Derivative Computations using CoDiPack.

@article{SaAlGauTOMS2019,
  title = {High-Performance Derivative Computations using CoDiPack},
  author = {M. Sagebaum, T. Albring, N.R. Gauger},
  url = {https://dl.acm.org/doi/abs/10.1145/3356900},
  year = {2019},
  date = {2019-12-01},
  journal = {ACM Transactions on Mathematical Software (TOMS)},
  volume = {45},
  number = {4},
}

About

Fast gradient evaluation in C++ based on Expression Templates.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • C++ 99.2%
  • Makefile 0.8%