aftershoq 1.0

A Flexible Tool for QCL/QCD Opitmization

aftershoq

A Flexible Tool for Em-Radiation-emitting Semiconductor Heterostructure Optimization using Quantum models

This Tool aims to aid in the simulation of quantum cascade structures (such as QC lasers, detectors, or QWIPs) using a variety of different simulation models. It also contains routines for optimization of such structures. It contains a libraty of common materials and structures used in the Litterature, and provides a framework for simulations. It does not contain any simualation code, this has to be provided by the users themselves (for now). The respective simulation code can be linked to [aftershoq] by the implementation of a subclass to Interface, which writes input files, executes the model, computes the merit function, and gathers the results data.

This is a program written for Python 3.6. You need to have Python 3 installed to use and modify this software to your needs. The current implentation also uses numpy, scipy, matplotlib, and lxml for some features.

Installation

When cloning, use the --recursive option:

git clone --recursive https://github.com/mfranckie/aftershoq.git

(or

git clone --recurse-submodules https://github.com/mfranckie/aftershoq.git

depending on your git version) so that the project "hilbert_curve" appears in the base directory of aftershoq. To install aftershoq and all its dependencies, execute

python setup.py install

from the aftershoq/ directory. To install on a system without root privileges, run

python setyp.py install --user

insead.

Tutorials

For a demonstration, see the Jupyter notebooks located in examples/notebooks. To install Jupyter, run

python -m pip install jupyter

then run with

jupyter notebook

Materials_guide.ipynb Shows how to create materials and alloys with varying composition and strain.

QCL_guide.ipynb Shows how to generate structures from scratch, how to load them from the library and how to generate them automatically.

Opt_guide.ipynb Shows how to setup and run an optimization with Gaussian Processs (GP) regression for a test function and for a real QCL (requires ownership of a separate QCL simulation model).

If you don't want to/can't use jupyter, the following examples have a similar content:

1. "QCLexample.py" (Requires a supported simulation program)
2. "example_sewself.py" (Requires the sewself program)
3. "example_sewlab.py" (Requires sewlab version 4.6.4 or later)
4. "test_optim.py" (No requirements, this is a test of the optimization scheme)

Good luck!