Companion repo for various articles:
-
[cutting2018] del Rosario, Fenrich, and Iaccarino, "Cutting the Double Loop: Theory and Algorithms for Reliability-Based Design Optimization with Statistical Uncertainty" (2018, Accepted) NME, pre-print
-
[beyond[2019] del Rosario, Fenrich, and Iaccarino, "Beyond Basis Values: Fast Precision Margin with FORM" (2019) 20th AIAA Non-Deterministic Approaches Conference
Core dependencies are Python with Numpy / Scipy and Matplotlib; I prefer to manage these dependencies through Anaconda Python. We also use a number of convenience functions available through a python package, which can be easily disabled if desired.
For the cantilever beam function, we use the Dakota implementation and optimizer.
For the fancy figures under /genfig, we use ggplot2, a package in R. These are extremely optional.
The repo is organized as follows:
./ # Root directory
./code/
./code/beam/ # Generates results for cantilever beam problem using MIP, from [cutting2018]
./code/genfig/ # Generates figures from saved data files, from [cutting2018]
./code/tension/ # Generates results for uniaxial tension problem, using MIP from [cutting2018]
./code/form/ # Generates results for cantilever beam problem using BIM, from [beyond2019]
./data/ # Output directory for data files
./images/ # Output directory for result images
Scripts under ./code/tension/ start with a Script parameters section, which can be used to explore different settings. The cases explored in the paper can be accessed by changing the MYCASE variable. These scripts are simply run from a terminal, e.g. python an_tension.py.
TODO Add instructions for ./code/beam/ files.
The .Rmd file in ./code/genfig/ is intended to be run from Rstudio, which aided in prototyping the figure aesthetics.