================================================
Kriging code for multidimensional interpolation
================================================
Main Author : Wataru Yamazaki
Modified Version: Komahan Boopathy,
University of Dayton, OH
komahan.cool@gmail.com
Date : 1/31/2014
==================================================
Article:
--------
K. Boopathy and M.P. Rumpfkeil,~\href{http://arc.aiaa.org/doi/abs/10.2514/1.J053064}{``Unified Framework for Training Point Selection and Error Estimation for Surrogate Models''}, AIAA
Journal, Vol. 53, No. 1, pp. 215--234, 2015, DOI: 10.2514/1.J053064.
Citation:
----------
@article{Komahan2013c,
author = "K. Boopathy and M. P. Rumpfkeil",
TITLE = "{Unified Framework for Training Point Selection and Error Estimation for Surrogate Models}",
journal = "AIAA Journal",
volume = "53, No. 1",
pages = "215--234",
doi = "http://dx.doi.org/10.2514/1.J053064",
year = "2015"
}
Main program can be executed within folders test/ or testdrag/ testlift/ as the case might be. The important .f90 files of interest will be:
main.f90 --> input parameters and settings (driver program)
monaco.f90 --> statistics and derivatives for UQ/OUU
post.f90 --> post processing on the surrogate (RMSE and other tecplot output coding)
DynamicPointSelection.f90 --> Dynamic training point selection based on local surrogate models
functions.f90 --> contains test functions, the user can add more.
Subroutine version can be evoked my running ./makelibrary in the main folder that creates krigingestimate.a for Optimization Under Uncertainty.
Dependencies:
-------------
- Tapenade for automatic differentiation
- MIR for local interpolation