summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
11 daysMore straightforward adjustment of step sizeJaron Kent-Dobias
11 daysStart μ largerJaron Kent-Dobias
11 daysIteration tweakJaron Kent-Dobias
11 daysBack to the old schemeJaron Kent-Dobias
11 daysDon't rely on past output for μJaron Kent-Dobias
11 daysSofter quitting requirements for μJaron Kent-Dobias
11 daysStart from β = 0 againJaron Kent-Dobias
11 daysBetter setting of μ boundsJaron Kent-Dobias
11 daysMore tweaksJaron Kent-Dobias
11 daysMore tweaksJaron Kent-Dobias
11 daysMake μ start a little smallerJaron Kent-Dobias
11 daysMake μ start a little smallJaron Kent-Dobias
11 daysImplemented binary search for μJaron Kent-Dobias
11 daysAnother sqrt?Jaron Kent-Dobias
11 daysMore tweaksJaron Kent-Dobias
11 daysMore daring with μμJaron Kent-Dobias
11 daysTweakJaron Kent-Dobias
11 daysTemper μ changesJaron Kent-Dobias
11 daysSlower μ stepsJaron Kent-Dobias
11 daysDon't adjust μ during the relaxation of C, RJaron Kent-Dobias
11 daysRaised minimum step sizeJaron Kent-Dobias
11 daysTweakJaron Kent-Dobias
11 daysMore tweaksJaron Kent-Dobias
11 daysTweaking step size adjustmentsJaron Kent-Dobias
11 daysSlow down μ correctlyJaron Kent-Dobias
11 daysAlso slow down μ changesJaron Kent-Dobias
11 daysLess feedbackJaron Kent-Dobias
11 daysMore feedbackJaron Kent-Dobias
11 daysPlaying with step size contralJaron Kent-Dobias
11 daysgamma to ΓJaron Kent-Dobias
11 daysChanged notationJaron Kent-Dobias
11 daysChanged interative scheme and many variable namesJaron Kent-Dobias
12 daysTweaking form of iterationsJaron Kent-Dobias
12 daysIgnore new binaryJaron Kent-Dobias
12 daysFixed MakefileJaron Kent-Dobias
12 daysRefactorJaron Kent-Dobias
12 daysFixed bug and got integrator workingJaron Kent-Dobias
12 daysGot the log-fourier method working, but integrator is not working correctlyJaron Kent-Dobias
13 daysStarted implementing log-Fourier method, not working right now.Jaron Kent-Dobias
2025-04-09Rewrote the original integrator to use a new approach.Jaron Kent-Dobias
2025-04-09Started making chages to the original integratorJaron Kent-Dobias
2025-04-05Made τ₀ = 0 possibleJaron Kent-Dobias
2025-04-05Switched to fftw allocators to ensure correct alignment for SIMD instructionsJaron Kent-Dobias
2025-04-05Enable multithreaded FFTW with openmpJaron Kent-Dobias
2025-04-05Ignore fftw wisdom.Jaron Kent-Dobias
2025-04-05Rescale τₘₐₓ with τ₀ to make approximations more consistent ↵Jaron Kent-Dobias
accross parameters
2025-04-05Save and read fftw wisdom to speed initializationJaron Kent-Dobias
2025-04-05Save half the data to the diskJaron Kent-Dobias
2025-04-05Slightly better memory efficiencyJaron Kent-Dobias
2025-04-04Adopted new scaling of Γ₀ to keep z and y moderateJaron Kent-Dobias