[Requirement] Remove low resolution ToF

[jl-wynen]

Executive summary

This method will be used to filter low resolution time-of-flight data (ToF).

Context and background knowledge

https://docs.mantidproject.org/v4.0.0/algorithms/RemoveLowResTOF-v1.html

(Note that the option using the Hodges criteria should not be implemented in this first version)

Inputs

• Scipp data array with ToF as coordinates
• Users should be able to specify either the minimum ToF or wavelength.

Methodology

Follow Mantid's implementation.

Outputs

Scipp data array with low resolution ToF cropped or masked

Which interfaces are required?

Python module / function, Python script

Test cases

csv file of DREAM mantle already used in Scipp documentation (see https://scipp.github.io/tutorials/from-tabular-data-to-binned-data.html)

Comments

No response

[SimonHeybrock]

@celinedurniak Given that the Hodges criterion should not be implemented, what is there to do here? Isn't this equivalent to setting a minimum wavelength?

[celinedurniak]

Yes it is for this first implementation. What I'd like to have is a function with a intuitive name so that users can apply this correction without knowing too much about Scipp / Python. Or will it be handle in a different way with Sciline?

[SimonHeybrock]

There will be a workflow that encapsulates this. The user can then set the TOF or wavelength range (similar example: https://scipp.github.io/esssans/examples/zoom.html#Setup-reduction-parameters, setting WavelengthBins).

Questions:

• Will this always (or most of the times) be set, i.e., should we include this as a fixed correction that is always enabled? In this case it could of course still be "disabled" by setting very wide bounds.
• Will it be either TOF or wavelength, or can it be both? Can we make a decision now, or should both be implemented?

[celinedurniak]

@SimonHeybrock Here are my answers to your questions. At the moment, I think

1. this algorithm will used most of the time
2. it is better to implement both: specifying ToF or wavelength

[SimonHeybrock]

Can you clarify if you mean that the Hodges criterion should be implemented? Or simply that we need to set bounds in ToF and wavelength? This is unclear from the current specification, since the Mantid algorithm uses the Hodges criterion when using ToF, as far as I can tell.

[jl-wynen]

@celinedurniak #44 has the option to crop ToF at the beginning of the workflow. It requires a manually-set parameter just like the POWGEN workflow: https://scipp.github.io/essdiffraction/examples/POWGEN_data_reduction.html#Define-reduction-parameters (ValidTofRange)

Is this sufficient? Or do we need to implement the Hodges criterion?

[celinedurniak]

@jl-wynen For the moment, the implementation of the Hodges criterion is not needed.

[jl-wynen]

Ok. Then I close this issue as completed.

[jl-wynen]

According to the PDF reduction documents shared by Celine (Mantid workflow), the step described here is performed after conversion to d-spacing and lorentz correction. There also is a separate 'crop in tof' step performed earlier after applying absorption corrections.

Found this in the part of the document that describes the basic powder reduction workflow. I'm reopening because we currently only have the latter.

[SimonHeybrock]

Can you elaborate why computing d-spacing or Lorentz correction matter? Neither should influence the TOF values, i.e., shouldn't the order of steps be irrelevant?

[jl-wynen]

@celinedurniak? I don't know. I'm assuming because those are needed to determine the resolution?