Next steps

[SimonHeybrock]

• [ ] Modify Sans2d workflow from esssans to return time-dependent $I_{\text{sample}}(\text{spin}, t, \lambda)$ in event mode with computed $Q_x$ and $Q_y$ (per event). @SimonHeybrock @nvaytet
  • The channel-, time-, and wavelength-binning must be explicit since we need to normalize to the corresponding channel-, time-, and wavelength bins of the monitor
  • Do we get result consistent with default workflow (replace polarization correction with identity matrix)?
• [ ] Implement functions in workflow skeleton (base.py)
• [x] Ultimately we plan to make esspolarization a dependency of esssans, which would combine the polarization-analysis workflow with the SANS workflow.
  • We will initially have a lot of changes and iteration in esspolarization. We can therefore consider depending on esssans in the docs of esspolarization (not in the package).
  • We can then add, e.g., SANS as an example in the docs of esspolarization.

[SimonHeybrock]

Regarding item 1.):

• If the monitor is not event-mode, we need to somehow split it along the time dimension. We could, e.g., assume that the monitor counts are proportional to the length of each time interval. Alternatively, one could use the relatively detector counts in each interval, which might be "safer" since it would take into account source current variations.
• For now, we can use the existing workflow without changes by requesting CleanQ[SampleRun, Numerator] and CleanQ[SampleRun, Denominator]. These are close to the final result, but without Q-binning and normalization, but otherwise exactly what we need (and the omitted steps are very simple).

[SimonHeybrock]

• If the monitor is not event-mode, we need to somehow split it along the time dimension. We could, e.g., assume that the monitor counts are proportional to the length of each time interval. Alternatively, one could use the relatively detector counts in each interval, which might be "safer" since it would take into account source current variations.

This monitor is positioned before all cells, i.e., the time-dependence of the transmission is irrelevant here. Everything should work as normal.

[SimonHeybrock]

See #24 on latest ideas regarding item 1.), closing as this is no longer fully relevant.