elkoa 1.3.0

Description

Elk Optics Analyzer (ElkOA) helps to analyze optics output data from The Elk Code.

Features

Elk Optics Analyzer…

• Comes with a GUI as well as a python CLI

• Supports Elk tasks 121, 187, 320 and 330

• Recognizes available tasks / (tensor) fields automatically

• Is easily extendable

Users can…

• Visualize real and imaginary parts of Elk optics output data in various ways

• Import additional data files, e.g. experimental measurements Ctrl+O

• Convert response functions via Universal Response Relations, e.g. ε ➙ σ Ctrl+C

• Convert dielectric tensors into (extra-)ordinary refractive indices for arbitrary wavevectors Ctrl+C

• Select tensor elements to plot individually via dialog Ctrl+T

• Use global tensor elements settings for all plots Ctrl+G

• Batch-load parameter studies to visually analyze the impact of different parameter settings Ctrl+B

• Write out displayed data in different formats Ctrl+W

Soon to come:

• 3D-plotting of index ellipsoids

• Batch-convert for a set of different q-points

• Sample/geometry-dependent (i.e. thin films) conversions of response functions

Requirements

• Python 3.x

• numpy

• matplotlib

• PyQt5

• pbr

You should use the packages provided by your linux distribution. On recent Debian systems for example, you can get all requirements by running

apt install python3-numpy python3-matplotlib python3-pyqt5 python3-pbr

Alternatively, you can get the latest PyPI versions of each package automatically by installing via pip (see below).

For testing purposes, you additionally need the following packages: * pytest * pytest-qt * pytest-mpl * nose

Installation

The easiest way to install ElkOA is via pip, either from PyPI directly

pip install elkoa

or, if you want the latest git version,

git clone https://github.com/PandaScience/ElkOpticsAnalyzer.git
cd ElkOpticsAnalyzer
pip install .

This will also install all required but absent python packages automatically from PyPI.

If you like to install ElkOA only for the current user, add the flag --user. If you want to take care of the required python packages yourself (i.e. by using the ones provided by your Linux distribution), add --no-deps. If you like to run a developer installation (no copying of files, instead use git repo files directly), add -e.

In any case, after installation you can run the ElkOA GUI from everywhere in a terminal using either elkoa or ElkOpticsAnalyzer.

Another way to install is by cloning the repo as above and instead of installing via pip, put something like

export PATH=$PATH:/path/to/ElkOpticsAnalyzer/elkoa/gui
export PYTHONPATH=$PYTHONPATH:/path/to/ElkOpticsAnalyzer/

to your .bashrc or .bash_profile. Then you can start the ElkOA GUI with ElkOpticsAnalyzer.py.

Tests

Testing is done using the pytest library. Make sure you installed all additional requirements beforehand.

1. Download and extract the sample data

   • TODO

2. Run (–mpl flag is mandatory!)

pytest test_figures.py --mpl

Python CLI

In an Elk output directory containing e.g. the files

elk.in INFO.OUT EPSILON_11.OUT EPSILON_12.OUT EPSILON_13.OUT EPSILON_21.OUT
EPSILON_22.OUT EPSILON_23.OUT EPSILON_31.OUT EPSILON_32.OUT EPSILON_33.OUT

you can run in a python3 interpreter:

from elkoa.utils import elk, io, convert
# parse Elk input file
elk_input = elk.ElkInput()
# read specific input parameter
eta = elk.readElkInputParameter("swidth")
# read tensorial Elk optics output (ij = dummy for 11, 12, etc.)
freqs, epsilon = io.readTenElk("EPSILON_TDDFT_ij.OUT")
# create converter instance
q = [0, 0, 0]
converter = convert.Converter(q, freqs, eta, opticalLimit=True)
# convert dielectric tensor to optical conductivity
sigma = converter.epsilonToSigma(epsilon)
# write out converted tensor
io.writeTensor("sigma_ij_test.dat", freqs, sigma, threeColumn=True)
# write out 11-element of converted tensor
io.writeScalar("sigma_11-scalar.dat", freqs, sigma[0, 0, :], threeColumn=True)

Misc

• Auto-converting filenames to tex-labels

  • For this feature to work, filenames must follow the pattern root+_sub+.ext, which will show up as rootsub.

  • In case root contains a case-insensitive substring like eps, EPSILON, Sig, SIGma etc., corresponding greek letters will be used, i.e. eps_ex.dat ➙ εex.

• Additional data plots

  • Number is restricted to 6, but in return we use consistent coloring after consecutively adding more plots.

Usage Examples GUI