Alexey Radul
Alexey Radul
There are several things we can do once #101 is merged that will make the resulting code base simpler. We should choose whether and when to. - [ ] Perhaps...
The first thing that one's eyes fall upon when visiting http://mitgcm.org/ are a set of colorful, reasonably high resolution graphics, presumably rendered from model results. This makes it instantly clear...
I am a little fuzzy on what exactly is involved in doing this. Some items that come to mind: - [ ] Claim the name `aronnax` on PyPI - [...
Said chatter would be - The progress reports called for by #13 - The start/stop notifications - Possibly some of the warnings (not errors) detailed in #45 Implementation options: -...
Process - [ ] Run a representative example under the profiler at low resolution - [ ] Fiddle with the parameters enough to characterize the scaling behavior of various parts...
Do the lunar and solar tides matter enough to model? How much implementation effort would it be to include them? What computational cost would they add?
Does the change in effective gravity produced by the Earth's rotation matter enough to model? How much implementation work would it be to include? What would its computational cost be?
MIM should check each of the invariants that it expects about its inputs, and emit a friendly error message if they are violated. The general principle is, don't allow the...
The benefits of single-precision are - Cross-checking: If the same simulation gives materially different results in single vs double precision, one can suspect severe accumulation of roundoff error. - However,...
The relevant dimensions of variability in computing environment are - Operating system - Operating system version - Fortran compiler - Fortran compiler version - Fortran compiler optimization flag setting -...