Optix

AeroLab Optimization Software

Created by:

• Dr Doug Hunsaker (professor, Utah State University; director, USU AeroLab)
• Josh Hodson (graduate student, Utah State University)
• Cory Goates (undergraduate/graduate student, Utah State University)

NOTE FOR LEGACY OPTIX USERS:

Since January of 2019, API of Optix has changed significantly to somewhat mirror that of scipy.optimize. We continue working to improve the API and functionality of Optix. This README will always contain up to date information for using Optix.

About

Optix is a gradient-based optimization tool designed with aerodynamics in mind. Recognizing that aerodynamic functions are often computationally intensive, Optix has been designed to be as light-weight and parallel as possible, while offering an intuitive API.

The algorithms have been developed, for the most part, using the following references:

• Parkinson, Balling, Hedengren, Optimization Methods for Engineering Design, Brigham Young University, 2013.
• Martins, Ning, Engineering Design Optimization, electronically published, 2020.

THIS SOFTWARE COMES WITH ABSOLUTELY NO WARRANTY EXPRESSED OR IMPLIED

API

All functionality of Optix is wrapped within the minimize(). The following is an example of how this function is used:

import optix as opt
from random import random

def f(x):
    return x[0]**4-2*x[1]*x[0]**2+x[1]**2+x[0]**2-2*x[0]+5

x0 = [-10+20*random(),-10+20*random()]

optimum = opt.minimize(f,x0,file_tag="_test",n_search=8,max_processes=8,line_search="quadratic",termination_tol=1e-6,verbose=False,hess_init=1)
print("Optimum value: {0}".format(optimum.f))
print("Optimum point: {0}".format(optimum.x))
print("Function calls: {0}".format(optimum.obj_calls))

File Outputs

Running optix.minimize() will also output results to 3 text files. These 3 files are named iterations, gradient, and evaluations, each appended with the user-specified file tag. The first two are written to during runtime. The last is written to only after successful completion of the optimization. Optimize mimics what is printed to the command line, giving information about the fitness, point in the design space, magnitude of steps, etc. Gradient gives information about the objective and constraint gradients at each point in the optimization. Evaluations simply outputs the value of the objective function at each point considered during optimization, including points used to calculate finite differences.

Documentation

For documentation, please refer to the docstrings. This can be done either by using the built in python help() command in the interpreter or by refering to the source. For example, to view the minimize() parameters and return values, type:

    help(optix.minimize)

into a Python interpreter. Please note that Optix should be imported first.

Getting the Source Code

The source code can be found at https://github.com/usuaero/Optix

You can either download the source as a ZIP file and extract the contents, or clone the Optix repository using Git. If your system does not already have a version of Git installed, you will not be able to use this second option unless you first download and install Git. If you are unsure, you can check by typing git --version into a command prompt.

Downloading source as a ZIP file

1. Open a web browser and navigate to https://github.com/usuaero/Optix
2. Make sure the Branch is set to Master
3. Click the Clone or download button
4. Select Download ZIP
5. Extract the downloaded ZIP file to a local directory on your machine

Cloning the Github repository

1. From the command prompt navigate to the directory where Optix will be installed
2. git clone https://github.com/usuaero/Optix

Installation

The Optix package can be installed by navigating to the root directory of the project and using the following command

   pip install .

This installs Optix as a package on your machine, meaning you can import Optix into any Python script without having to copy around the source files

Testing

Unit tests are are run using the following command:

    py.test

Please note that you must have Pytest installed for this to work.

Support

For bugs, please create an issue on the github repository.

License

This project is licensed under the MIT license. See LICENSE file for more information.