search

bradkav / DirectDetectionStandard

Repo and wiki for coming up with a set of standard inputs and outputs for dark matter direct detection codes
0 stars 0 forks source link

Levels of detail #3

Open bradkav opened 7 years ago

bradkav commented 7 years ago

Do we want to have multiple levels of detail/abstraction?

For a detailed analysis, you probably want to know the observed and background rates as a function of energy, detailed information about energy resolutions etc. But that obviously requires a bit more work - from the point of view of experiments providing information and from the point of view of theorists analysis the 'data'.

In some cases, you can do a rough job with (say) the number of observed events, the number of expected background events and an efficiency curve.

Should we allow for these two different levels of detail?

Alternatively, we can always calculate the 'coarse' information (number of observed events, number of expected events etc.) from the detailed 'fine-grained' information provided by the experiments.

jaydenn commented 7 years ago

I think we should allow for the two levels of detail, this is what I tried to do with the format of the LUCKS_runs.dat file. It would be great if this file was the bare minimum which could be provided. Missing resolution data could mean we assume perfect resolution, and missing efficiency data means we assume 100% efficiency. I am not sure what we can assume about the background, perhaps we should add a column for total expected background?

bradkav commented 7 years ago

Currently we have a column in LUCKS_runs.dat for N_obs, as well as a file which lists the events. So I guess we could have a column for expected background events N_BG in addition to the file with the background spectra.

It would be good then if we came up with a list of what the bare minimum information requirement is (say N_obs, N_BG, exposure, E_min, E_max, and maybe others) - as well as what the default assumptions are if a piece of information is missing.