Neural Network Parametrization of

 Deep-Inelastic Structure Functions

 
       This webpage provides downloads of the fortran code for a neural network parametrization of deep-inelastic proton, deuteron and non-singlet structure functions ready for applications, based on all available experimental data.

The parametrization of the proton structure functions is presented in the paper:

"Unbiased determination of the proton structure function F2p
with faithful uncertainty estimation"

  The NNPDF Collaboration:
Luigi Del Debbio, Stefano Forte, Jose Ignacio Latorre, Andrea Piccione, Joan Rojo

  http://arxiv.org/abs/hep-ph/0501067

"Neural network parametrization
of Deep-Inelastic Structure Functions"
 
Stefano Forte, Lluis Garrido, Jose Ignacio Latorre, Andrea Piccione.

  http://arxiv.org/abs/hep-ph/0204232
where a full presentation of this work can be found.

All-in-one download

• Single download of all programs and help: f2neural.tgz

Fortran code and weights

• fortran code: f2.f
• neural network weights (read by f2.f): f2p.weights, f2d.weights, f2ns.weights

Help for individual subroutines

• NET (man page: net.txt)
  Calculates the value of the structure function F2 from a specific neural network.
• F2AV (man page: f2av.txt)
  Calculates the value of the structure function F2 and its error for a given point.
• F2COR (man page: f2cor.txt)
  Calculates the value of the structure function F2 for two points with errors and correlation.

Test

• An example of how to insert the above fortran code is illustrated in test.f

Numerical results and plots

Go to on-line numerical results or on-line plots.




                                   
                              Comments to: nnpdf (at) mi.infn.it