Determine default for sampler, sampler settings, jump proposals

This work is already underway, but I'm creating a git issue to link to the review readiness wiki.

@gregory.ashton, @christopher-berry, @paul-lasky, @vivien

changed the description

added Sampling Testing + 1 deleted label

removed 1 deleted label

changed milestone to %1.0.0

added 1 deleted label

added 1 deleted label

assigned to @christopher-berry

Currently, default samplers might be:

• CPNest
• PTMCMC

as standard "Cloud-based" and "Chain-based" samplers. Subject to change and setting specifications with the work on comparing runs and assessing correctness of results in #306 and #248 (closed) .

Marking this as closed as the defaults have been set the recommended values (in both bilby and bilby_pipe).

closed

reopened

I want to leave this open until we have demonstrated all samplers to be working properly across parameter space.

Hi @christopher-berry, as discussed by email I think this is an issue which requires a long-term effort. As such, I'll move it to the feature requests wiki.

I'll also add that demonstrating all samplers (we have 8 I think) work across the parameter space is going to be a gigantic challenge, I predict we will never have the available people power to do this. If you have a specific task in mind, could I suggest this remains closed and we instead open an issue "Run sample XXX using the bilby_pipe_review script" as a new issue? For info on the review script, see, here.

closed

unassigned @christopher-berry