Open mdietze opened 9 years ago
@bcow and @ashiklom made a first draft of a Monte Carlo global sensitivity analysis here: https://github.com/bcow/Global-sensitivity-analysis
In an email conversation with @xuchongang, he has interest in adding FAST to the set of global SA supported by PEcAn. The current plan is to merge in the draft Monte Carlo SA by @bcow and @ashiklom this summer and then pass that (and the sampler code by @istfer #778) to Chonggang as an example that he can then generalize.
This issue is stale because it has been open 365 days with no activity.
The Sensitivity and Ensemble analysis code should be modified to allow selection and computation of alternative indices (eg Sobol) and alternative ensemble samplers (eg LHC, joint random, Sobol, etc). This will make it easier to perform global SA (using the Ensemble Analysis) and to use the EA & SA as the initial round of emulation.