.. _04_dft_calculations:

.. include:: ../../examples/04_dft_calculations/README.rst

----

.. code-block:: python

   from eminus import Atoms, SCF
   

Start by creating an :code:`Atoms` object for helium

Use a very small :code:`ecut` for a fast calculation

.. code-block:: python

   atoms = Atoms("He", [0, 0, 0], ecut=5)
   

Optional parameters with examples are listed as follows

Dictionary to set the maximum amount of steps per minimization method and their order

Set it to a very small value for a short output

.. code-block:: python

   opt = {"pccg": 5}
   

The :code:`SCF` class only needs an :code:`Atoms` object, but only calculate 5 steps for less output

.. code-block:: python

   print("First calculation:")
   SCF(atoms, opt=opt).run()
   

Exchange-correlation functional description (case insensitive), separated by a comma

.. code-block:: python

   xc = "lda,pw"
   

The Libxc interface can be used by adding :code:`libxc:` before a functional (you can also neglect the :code:`libxc` part)

Names and numbers can be used, and mixed with the internal functionals as well

.. code-block:: python

   # xc = "libxc:LDA_X,:LDA_C_PW"
   # xc = "libxc:1,pw"
   

Type of potential (case insensitive)

.. code-block:: python

   pot = "gth"
   

Initial guess method for the basis functions (case insensitive)

Adding :code:`sym` to the string will use the same guess for all spin channels

An unsymmetric guess will be used by default

.. code-block:: python

   guess = "random"
   

Convergence tolerance of the total energy

.. code-block:: python

   etol = 1e-8
   

Convergence tolerance of the gradient norm

.. code-block:: python

   gradtol = 1e-7
   

Calculate a self-interaction correction from the Kohn-Sham orbitals after an SCF calculation.

.. code-block:: python

   sic = True
   

Calculate a dispersion correction after an SCF calculation (need the dispersion extra).

.. code-block:: python

   disp = False
   

The amount of output can be controlled with the verbosity level

By default the verbosity level of the :code:`Atoms` object will be used

.. code-block:: python

   verbose = 4
   

Start a new calculation with new parameters

.. code-block:: python

   print("\nSecond calculation with more output:")
   scf = SCF(
       atoms=atoms,
       xc=xc,
       pot=pot,
       guess=guess,
       etol=etol,
       gradtol=gradtol,
       opt=opt,
       sic=sic,
       disp=disp,
       verbose=verbose,
   )
   

Arguments for the minimizer can be passed through via the run function, e.g., for the conjugated-gradient form

.. code-block:: python

   etot = scf.run(cgform=2)
   

The total energy is a return value of the :code:`SCF.run` function, but it is saved in the :code:`SCF` object as well with all energy contributions

.. code-block:: python

   print(f"\nEnergy from SCF function = {etot} Eh")
   print(f"\nEnergy in Atoms object:\n{scf.energies}")

Download :download:`04_dft_calculations.py <../../examples/04_dft_calculations/04_dft_calculations.py>`