ONTIME, LIVETIME, EXPOSURE, DTCOR in merged products

[kglotfelty]

Outputs from merge_obs and combine_spectra are doing {stuff} to provide meaningful values for some subset of the ONTIME, EXPOSURE, and LIVETIME keywords.

This issue is to discuss which keywords are and should be being updated and how. For example

• Should we be writing ONTIME as SUM(ONTIME_n)
• Should we be writing DTCOR as the ONTIME weighted sum of the individual DTCOR_n values
• Should we go ahead and delete per-chip values? Do they still have any meaning in the merged products?

Some may be tool specific (eg combine_spectra with method=avg|sum may have different behavior).

[DougBurke]

I think I have been using the @ gti trick to ensure we filter things like the asol file to match the data (memory is fuzzy and vague) and I know that appears to be problematic (I didn't follow the details so it hopefully isn't an issue for what we are doing). if we are doing it.

[kglotfelty]

The @ gti thing is only a problem if you filter and bin in the same command.

So if a user supplies fluximage an infile with "evt[@ gti]" then they might not be getting the correct image. But if just using to filter a table then it looked to be OK.

[kglotfelty]

The @ gti (no space) was tagging someone

[kglotfelty]

Note: this looks to be a duplicate of #177 ?