Workflow capabilities list

[jokasimr]

This is a list of the things that reflectometry users might want to do:

• Create a reference $I_{ref}(px, t)$ corrected by the expected reflectivity of the supermirror.
  • It does not have to be represented in the coordinates pixel and event_time_offset. Instead of pixel we could have some other coordinate like $\delta$ (the angle offset from the center of the detector), as long as the coordinate don't depend on instrument settings that vary between the reference and sample measurement. It's also probably better to use $\lambda$ instead of $t$. Done.
  • Maybe this file can be cached on disk, that's what's done in the amor code. The reference measurement might be quite large (about 5x a sample measurement from what I understand). But I don't think caching it on disk is necessary, it's probably enough to cache it in memory when treating multiple sample files, and sciline does that automatically. Done.
    *It should be possible to compute reflectivity ($R$) as a function of $Q$ or $Q_x, Q_y$ or as a function of $\delta, \lambda$. Created issue, see [Requirement] Option to compute $R(Q_x, Q_y)$  #61.
• We need to be able to do time-resolved experiments. Issue created, see [Requirement] Support for time resolved experiments #57
• Should be simple to display several $R(Q)$ curves on top of each other. For example, runs with different $\mu$ (translates to different $Q$-range). Or, runs with different samples. Done
• Should be simple to treat several files as a single file with a larger event stream.
• The sample events should be filtered based on $x$-range (the off-specular scattering direction), and on wavelength, and on $Q$ (?), and based on reference measurement counts (if the reference measurement has very high uncertainty in a region then we should remove counts in that region (?)). Done
• Footprint correction should be available but not applied in most cases. For ESTIA the footprint correction can be done in hardware by adjusting some slit size (?) but as I understand it we still need the option in case of particularly small samples. Done / not relevant yet, revisit when the ESTIA workflow is specified.
• The final result $R(Q)$ should be re-scaled so that the reflectivity is 1 for low $Q$. (Average $R(Q)$ over a low-$Q$ interval and scale by that value). Issue added [Amor only] Scale R(Q) so that reflectivity is 1 for low Q #59
• (maybe) Correct L1 by the focusing guides. The focusing machinery makes the path from chopper to sample non-straight, the current code assumes a straight path. This effect is small, but it should be visible once we have corrected for other issues. Related to [Amor] Make results match Jochens data reduction program #60

[jokasimr]

Created separate issues for the items here. Closing.