Closed ojscutt closed 5 months ago
Plot for n=6 to 40, hare taken from the grid:
Plot for n=15 to 25, hare from grid:
Plot for n=6 to 40, emu hare:
Plot for n=15 to 25, emu hare:
The sampler spins out here either way - things seem to behave when we include all mode frequencies, but not when we restrict the constraints to just n=15 to 25, regardless of whether the Hare is taken from the grid or the emulator.
Solved by using ultranest instead of dynesty - likelihood, emulator, seem to be working as expected.
Plot(s) here of comparison between sampling runs with grid hare and emulator hare: