README.md

Our publication is in review, but you can have a look at our preprint in the meantime!

Contents

0. Intro to Python (for R programmers)
1. Fundamentals
   1. Quickstart
   2. Workflow Breakdown
2. Landscape
   1. Basic Landscape Elements
   2. Heterogeneous Point-Types
   3. Custom Migration Matrix
3. Optimization
   1. Genetic Algorithm Primer
   2. Discrete Optimization (GA)
   3. Continuous Optimization (GA)
4. Analysis
   1. Visualization Tools
   2. GA Comparison
5. Advanced Topics
   1. Mathematical Formulation
   2. OpenStreetMap Compatibility
   3. Particle-Swarm Optimization (PSO)
6. Preprint's Demos
   1. Preprint on bioRxiv
   2. São Tomé
   3. Yorkeys Knob

Installation Instructions

The materials for this webinar can be run by installing the codebase locally, or by using our docker image. The easiest way to get all the materials for the webinar is to install Docker and download our container with the following command in our terminal:

docker pull chipdelmal/mgsurve_webinar2023

Once the image is downloaded, we can run the contents with:

docker run -p 8888:8888 chipdelmal/mgsurve_webinar2023

and head to the following address in our internet browser: http://127.0.0.1:8888/lab; where a JupyterLab session should be running with all the required contents and packages for the course. To open up the webinar's index, open the README.md file in the jupyterlab session, right click anywhere in the document, and select Show Markdown Preview.

The changes we make within the Docker container won't be saved across sessions. If we want to make take notes on the notebooks and make changes session-persistent, first run:

docker volume create mgsurve

to create a docker volume. We can then launch the image with the volume attached with the following command:

docker run \
   -v mgsurve:/MGSurvE_Webinar2023 \
   -p 8888:8888 chipdelmal/mgsurve_webinar2023 

Additional instructions for local installation are provided, as well as a more thorough set of alternatives in our package documentation in case the Docker alternative is not suitable for any attendee!

Video Recording

To watch the recording of the webinar session click on our banner!

Authors and Funders

Dev and Webinar: Héctor M. Sánchez C.

PIs: John M. Marshall, David L. Smith

   

Special Thanks

• Tomás León for allowing us the use of his São Tomé migration matrix, and for advice in the ecology of the application.
• Jared B Bennett for tips and tricks on software and math for the optimization process.
• Lillian Weng, Xingli Yu, Ayden Salazar, Topiltzin Hernández Mares, and Joanna Yoo, for their help in developing the PSO algorithm.
• Elijah Bartolome for several benchmarks on the speed and accuracy of different spherical distance functions.