search

claudiatebaldi / hectorcal

3 stars 3 forks source link

Support for MCMC calculation #1

Closed rplzzz closed 5 years ago

rplzzz commented 5 years ago

This branch includes the krd_single_ESM_calibration branch, so I've targeted the PR there to make it easier to see what's changed. I think what we should do is to wait to merge until the single-ESM branch is ready to go, then do the merges all at once.

rplzzz commented 5 years ago

Since this never got merged, I'm going to continue developing the latest mods to the likelihood function on it and mark it ready to merge once we have it working with the new system.

rplzzz commented 5 years ago

Good news, everyone. As of now I am reasonably confident that the Bayesian MC functions are now working correctly with the ocean heat flux. That means that the next step is to try running some Monte Carlo runs to see what we get, which is to say that if fortune smiles upon us we should be pretty close to closing out the calculations for Paper II. Huzzah!

It looks like some conflicts have crept in between my branch and the singleESM_calibration branch, so maybe I'll try to fix those tomorrow, and then maybe try to cobble together some MC calculations in between the sessions at the workshop next week. In the meantime, I think I'll take the rest of the evening off to watch a kung fu movie or something.

rplzzz commented 5 years ago

@kdorheim Since the package infrastructure for the MCMC is complete, I think we should merge this in with your development branch so as to minimize the chances of any further branch divergence. If you would, please, take a look at it sometime during the upcoming week and see if you agree.

rplzzz commented 5 years ago

@kdorheim Any objection to me merging this?