Fortformat
[Input: recipes/fortformat/]
This chapter should serve as a tutorial guiding you through your first dataset
creation and data extraction after running Fortnet by using two Python classes,
provided by the Fortformat Python package:
After these tutorials, you will be able to create a Fortnet compatible dataset, based on the output files of your simulation package of choice (e.g. VASP), as well as extracting the resulting output, i.e. predictions of the network.
Fortformat: Basic Fortnet IO Format Classes
This Python package provides two basic classes that implement the associated Fortnet input and output file format. The input features and targets, stored in lists of Numpy arrays, conveniently get dumped to disk as HDF5 files, while Fortnet’s output may be extracted from the HDF5 output files as written by a validation or prediction run.
Installation
Please note, that this package has been tested for Python 3.X support. It additionally needs numerical Python (the Numpy module) as well as h5py.
Since the Fnetdata class, among others, expects the so-called Atoms
objects of the Atomic Simulation Environment
(ASE) as an input, sooner or later this
dependency will also have to be satisfied.
System install
You can install the script package via the standard Python setup mechanism. If you want to install it system-wide into your normal Python installation, you simply issue
python setup.py install
at tools/fortformat/ with an appropriate level of permission.
Local install
Alternatively, you can install it locally in your home space, e.g.:
python setup.py install --user
If the local Python install directory is not in your path, you should add this. For the bash shell you should include the following line in the .bashrc:
export PATH=$PATH:/home/user/.local/bin