summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2020-11-01Added correlation code.Jaron Kent-Dobias
2020-06-26Used the wrong variable for number of tuning sweeps for step.Jaron Kent-Dobias
2020-06-26Added ability to load temperatures from file.Jaron Kent-Dobias
2020-05-12New variable in tuning that sets how much of new temperatures to use.Jaron Kent-Dobias
2020-05-12Conditional failure.Jaron Kent-Dobias
2020-05-12Print more often.Jaron Kent-Dobias
2020-05-12Don't stop tuning if we haven't had full transits.Jaron Kent-Dobias
2020-05-12Testing new strategy for judging tuning convergence.Jaron Kent-Dobias
2020-05-12Added more debug printing while tuning.Jaron Kent-Dobias
2020-05-12Removed overengineered tuning system.Jaron Kent-Dobias
2020-05-12Added new code for measurements, and utility for investigating the Kauzmann ↵Jaron Kent-Dobias
paradox.
2020-05-12Constructor now allows setting θ₀ manually.Jaron Kent-Dobias
2020-05-12Step angles drawn from Gaussian rather than uniform distribution.Jaron Kent-Dobias
2020-05-12Two changes to MCMC library.Jaron Kent-Dobias
- Added orthogonalize command that resets matrix to a nearby orthogonal one. - run command now returns proportion of acceptances.
2020-03-17Interactive choice for whether to finish tuning.Jaron Kent-Dobias
2020-02-05removed unused range typeJaron Kent-Dobias
2020-02-03Ran clang formatJaron Kent-Dobias
2020-02-03Givens rotations now use single vector for storage instead of allocating one ↵Jaron Kent-Dobias
each construction
2020-02-03replace openmp parallelization with std::foreachJaron Kent-Dobias
2020-02-03replaced mersenne twister with pcg32Jaron Kent-Dobias
2020-01-30any orthogonal size that is a multiple of four is now available instead of ↵Jaron Kent-Dobias
powers of two
2019-12-22fewer pt tries per sweepJaron Kent-Dobias
2019-12-21now using random choice of pairs to swap instead of sweeping left to rightJaron Kent-Dobias
2019-12-21more gentle increase of tuningJaron Kent-Dobias
2019-12-20fixed itJaron Kent-Dobias
2019-12-20testingJaron Kent-Dobias
2019-12-20maybe nowJaron Kent-Dobias
2019-12-20maybe fixed itJaron Kent-Dobias
2019-12-20another tryJaron Kent-Dobias
2019-12-20another tryJaron Kent-Dobias
2019-12-20another tryJaron Kent-Dobias
2019-12-20tweaksJaron Kent-Dobias
2019-12-20tweaksJaron Kent-Dobias
2019-12-20maybe fixed nonconstant problemJaron Kent-Dobias
2019-12-20trying to remove anamolous acceptance rates from edgesJaron Kent-Dobias
2019-12-19new conditions for stopping tuningJaron Kent-Dobias
2019-12-19command line flag for temperature convergence precisionJaron Kent-Dobias
2019-12-19working pt with tuningJaron Kent-Dobias
2019-12-18tuning now progresses until convergenceJaron Kent-Dobias
2019-12-18half implemented new tuning for ptJaron Kent-Dobias
2019-12-17partial implementation of dynamic ptJaron Kent-Dobias
2019-12-12bug fix: parallel measurements recorded during dry runsJaron Kent-Dobias
2019-12-12added a stage for burnout steps to address superheatingJaron Kent-Dobias
2019-12-11ran clang-formatJaron Kent-Dobias
2019-12-11ranges are no longer hardcodedJaron Kent-Dobias
2019-12-11some support for the denser temperatures around the transition in the main ↵Jaron Kent-Dobias
file, but values are currently hardcoded in
2019-12-11added support in parallel tempering for specifying different densities of ↵Jaron Kent-Dobias
temperatures in different intervals
2019-12-10implemented measurement of disorder parameterJaron Kent-Dobias
2019-12-10mistake in initialization led to incorrect energiesJaron Kent-Dobias
2019-12-10cleanup, and change to initalization with walsh matricesJaron Kent-Dobias