SANN_NN.py - SANN network classification
========================================

REQUIREMENTS
------------

Python 2 or 3 with numpy and scipy libraries installed.


HOW TO RUN THE PROGRAM
----------------------

In order to run the program the following files must be in the same
folder: SANN_NN.py, particles.py, a file with extension xyz

To run the program:

python SANN_NN.py  filename.xyz  size_ratio  pbc

where:

- filename.xyz is the name of the file with the snapshot to classify.
  Must have xyz extension.

- size ratio can be 0.45, 0.54, 0.82 and specifies the neural network to use.

- pbc can be either 0 or 1. Choose 1 to apply periodic boundary conditions,
  0 otherwise.


XYZ FILE FORMAT
---------------

The xyz file containing the snapshot must be formatted as follows:

<number of particles>
<Lx Ly Lz>
<label> <x> <y> <z> <diameter>
<label> <x> <y> <z> <diameter>
...

Specifications:

- Lx, Ly, Lz in the second line are the linear sizes of the box in the
  x, y and z directions. They are necessary if periodic boundary conditions
  must be applied, otherwise this line can be left empty (but must exist).

- label is simply a label associated to the particle. It can be a float 
  or an integer. It does not need to have any particular meaning in the input file.

- x, y and z are the coordinates of the particle.

- diameter is the diameter of the particle. The value itself is not important,
  BUT it is important that the larger species have a larger diameter.


OUTPUT OF THE PROGRAM
---------------------

The program produces an output file with the same name as the input file
+ "_results" and an xyx extension.
This file is the same as the input file, where the label of each particle is
substituted with the result of the SANN network classification as described in the paper.