https://github.com/perdumonocle/neuralnet2cpp

Compute your R 'neuralnet' neural network using cpp code
https://github.com/perdumonocle/neuralnet2cpp

c code-generation code-generator codegen codegenerator compute converter cpp neural-network neuralnet r

Last synced: about 1 month ago
JSON representation

Compute your R 'neuralnet' neural network using cpp code

• Host: GitHub
• URL: https://github.com/perdumonocle/neuralnet2cpp
• Owner: perdumonocle
• License: mit
• Created: 2017-06-12T19:37:03.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2017-06-13T06:39:50.000Z (almost 8 years ago)
• Last Synced: 2025-01-22T11:47:05.705Z (3 months ago)
• Topics: c, code-generation, code-generator, codegen, codegenerator, compute, converter, cpp, neural-network, neuralnet, r
• Language: R
• Size: 10.7 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# neuralnet2cpp

Compute your R 'neuralnet' neural network using cpp code.

You can train you neural network with R:


> library( 'neuralnet' )

> data <- data.frame( i1 = c(1, 2, 4), i2 = c(5, 5, 6), o = c(1, 1, 0) )

> net <- neuralnet( o~i1+i2, hidden = c( 5, 2 ) )



Then test your net:


> test <- data.frame( i1 = 5, i2 = 1 )

> res <- compute( net, test )

> res$net.result

               [,1]

[1,] -0.08056605515



And convert that network into cpp:


> compute.gen.cpp( net, "~/myprojects/subfolder/compute.cpp" )



Test your network using compute.cpp:


int main(int argc, char *argv[])

{

    const double covariates[2] = { 5, 1 };

    double results[1] = { 0 };

    neuralnet::compute( &covariates[0], &results[0] );

    printf( "%f\n", results[0] ); // -0.080566

    return 0;

}