AutoDiff .NET
Summary:
A simple .NET library for evaluating the value/gradient of a function using reverse-mode automatic differentiation.
URL: http://autodiff.codeplex.com
Developers:
- Alex Shtof
| Mode: |
Reverse |
| Method: |
Operator overloading |
| Supported Language: |
.NET |
Features:
* A pure .NET library that works on all .NET platform languages.
* Supports basic arithmetic operations, power, exponential and
logarithm functions
* Allows user-definable functions that are not supported by the
"core" library (using delegates).
* Supports basic vector math to support easier definition of
functions that work with vectors
* Function value and gradient evaluation.
* Uses the new .NET code contracts for parameter validation
(means
better performance and documentation)
* Works on Windows and on Linux (with Mono 2.10 and later)
Supported Platforms:
- Windows
- Unix/Linux
Licensing: open source
Entries in our publication database that actually use AutoDiff .NET in the numerical experiments: 0
The following diagram shows these entries versus the year of the publication.
|
|||
| Year | |||