This issue is just a to-do list of what we need to do to wrap up the analysis of the challenge:
[x] Run fiducial runtime and signal to noise tests, get agreement between 2-3 people.
[x] Run test of scaling of time vs number of auto- and cross-corrs, get agreement between 2 people.
[ ] Run test of scaling of time vs required delta chi^2, get agreement between 2 people (prereq still needed: accuracy settings from Levin and FKEM entries.)
[x] Finalise benchmarks for quarter and half bin width.
[x] Run test of scaling of accuracy vs bin width, get agreement between 2 people.
[x] Run test of scaling of time vs number of cores / nodes (parallelisability test).
[ ] Figure out why Tassia is getting a different Levin runtime from Danielle / David.
[ ] Iterate with entry authors for any last tweaks following the results of these tests.
This issue is just a to-do list of what we need to do to wrap up the analysis of the challenge: