.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/gap_fit.sh"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

    .. note::
        :class: sphx-glr-download-link-note

        :ref:`Go to the end <sphx_glr_download_auto_examples_gap_fit.sh>`
        to download the full example code.

.. rst-class:: sphx-glr-example-title

.. _sphx_glr_auto_examples_gap_fit.sh:

Fitting GAP 
===========

In this tutorial, you will learn to write a bash-script that fits GAP on your DFT data.
GAP learns the energies and forces as a function of the atomic positions. 
It is implemented as a command line tool ``gap_fit``
which takes many parameters. 
The parameters are specified in the following, and then passed to ``gap_fit``

.. GENERATED FROM PYTHON SOURCE LINES 9-12

First, we define our parameters which we want to vary later
In this case, that's the Cut-off of the atomic neighborhood,
the output file and the input file.

.. GENERATED FROM PYTHON SOURCE LINES 14-17

.. code-block:: Bash

    CUT_OFF=4.0
    GAP_FILE=cut_off_4A/SOAP.xml
    INPUT_FILE=train.xyz

.. GENERATED FROM PYTHON SOURCE LINES 18-19

Next, we have the parameters for GAP. 

.. GENERATED FROM PYTHON SOURCE LINES 20-20

.. code-block:: Bash

    GAP_PARAMS=(
.. GENERATED FROM PYTHON SOURCE LINES 21-22

First, we have the parameters for the 2 body distance descriptor

.. GENERATED FROM PYTHON SOURCE LINES 23-31

.. code-block:: Bash

    distance_2b
    cutoff=$CUT_OFF
    covariance_type=ard_se
    delta=1
    theta_uniform=1.0
    sparse_method=uniform
    n_sparse=300
    Z1=18
    Z2=18 :
.. GENERATED FROM PYTHON SOURCE LINES 32-33

Then we have the parameters for the SOAP descriptor

.. GENERATED FROM PYTHON SOURCE LINES 34-34

.. code-block:: Bash

    soap 
.. GENERATED FROM PYTHON SOURCE LINES 35-36

``n_max`` and ``l_max`` is the order of expansion in spherical harmonics

.. GENERATED FROM PYTHON SOURCE LINES 37-38

.. code-block:: Bash

    l_max=6
    n_max=6
.. GENERATED FROM PYTHON SOURCE LINES 39-40

``atom_sigma`` is the smearing of the atomic position

.. GENERATED FROM PYTHON SOURCE LINES 41-42

.. code-block:: Bash

    atom_sigma=0.5
    zeta=4
.. GENERATED FROM PYTHON SOURCE LINES 43-44

Then we specify the cut-off for the neighborhood around the centered atom

.. GENERATED FROM PYTHON SOURCE LINES 45-46

.. code-block:: Bash

    cutoff=$CUT_OFF
    cutoff_transition_width=0.5
.. GENERATED FROM PYTHON SOURCE LINES 47-48

and some magic settings

.. GENERATED FROM PYTHON SOURCE LINES 49-52

.. code-block:: Bash

    covariance_type=dot_product
    n_sparse=300
    sparse_method=random
    delta=1.0
.. GENERATED FROM PYTHON SOURCE LINES 53-54

Finally, we specify the number of different species and the atomic charge number of the species

.. GENERATED FROM PYTHON SOURCE LINES 55-58

.. code-block:: Bash

    n_Z=1 Z={18}

    )

.. GENERATED FROM PYTHON SOURCE LINES 59-60

Next, we have some general parameters

.. GENERATED FROM PYTHON SOURCE LINES 61-63

.. code-block:: Bash

    GEN_PARAMS=(
    e0_method=average 
    gap={"${GAP_PARAMS[@]}"}
.. GENERATED FROM PYTHON SOURCE LINES 64-65

Here we specify the input file (the atomic coordinates and forces) and the output file (the machine learned potential)

.. GENERATED FROM PYTHON SOURCE LINES 66-69

.. code-block:: Bash

    at_file=$INPUT_FILE
    gp_file=$GAP_FILE 
    default_sigma={0.003 0.15 0 0}
    sparse_jitter=1.0e-10
.. GENERATED FROM PYTHON SOURCE LINES 70-71

and then we have some parsing options for the input file

.. GENERATED FROM PYTHON SOURCE LINES 72-75

.. code-block:: Bash

    force_parameter_name=forces
    energy_parameter_name=energy
    )

.. GENERATED FROM PYTHON SOURCE LINES 76-77

Finally, we print our parameters and then run the gap_fit command

.. GENERATED FROM PYTHON SOURCE LINES 78-81

.. code-block:: Bash

    echo ${GEN_PARAMS[@]}

    gap_fit "${GEN_PARAMS[@]}"

.. GENERATED FROM PYTHON SOURCE LINES 81-83

To run this script open the terminal, navigate to the folder ``your_project/gap`` 
and execute the command ``bash gap_fit.sh``. 


.. _sphx_glr_download_auto_examples_gap_fit.sh:

.. only:: html

  .. container:: sphx-glr-footer sphx-glr-footer-example

    .. container:: sphx-glr-download sphx-glr-download-python

      :download:`Download Bash source code: gap_fit.sh <gap_fit.sh>`

    .. container:: sphx-glr-download sphx-glr-download-zip

      :download:`Download zipped: gap_fit.zip <gap_fit.zip>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_