Items pulled from prior Roadmap discussions

[butlerpd]

As of July 24, 2022, the following items, raised at prior roadmap discussions, don't seem to have ever made it to being integrated to any roadmap. Some may in fact be in different ways. Some may not be but are now obsolete for other reasons etc. I have included some notes where I'm unsure. Those that were obviously included, obsolete or already fixed were not transferred here in the first place.

• Choose better graphics package vs matplotlib? (qtplot, pyqtgraph,Silex, other).. or maybe extra for certain tasks?
• Quick start guides for adding models and scripting (Paul's note: is this still the best if we have both a written and youtube tutorials?)
• Finish SASfit model integration (Paul's note: I think this is done?)(Steve's note: So as we discussed on the call, how close are we to being able to take a SASfit model and use it as a plugin?)
• Custom workflows pol beam, contrast series, etc
• Inequality constraints -and work with polydispersity
• Add custom cost functions (other than xi^2 ... such as proportional to Q etc)
• Gen Scattering calculator -- does it include H exchange? What about missing H in the PDB? (Paul's note: Not sure this is something that should be in the 5 year roadmap at this point?)
• Sorting of chain fitting in batch (Paul's note: is this already done?)(Steve's note: Is it this: https://www.sasview.org/docs/user/qtgui/Perspectives/Fitting/fitting_help.html?highlight=chain+fitting#combined-batch-fit-mode) (Paul's note 2: No. This is about being able to change the order of the data being fit so that the chain fitting -using the result from one fit as the starting parameters for the next- is useful. Originally the list was in random order which makes chain fitting a bit pointless). TESTING OF 5.0.5 SHOWS THIS IS DONE. checking "chain fit" adds a tab called "order" allowing one to order the datasets.
• Complex sld (Paul'ls note: how much of this is done already?)
• Qz support (Paul's note: Also highly requested by mag SANS workshop)
• Supply test data used in sasmodel documentation png file (Paul's note: Is that really necessary)
• Naryanan == ASAXS is niche should be very low priority (Paul's note: should we remove this from the roadmap until there is demand? --> ask Snow?)
• Naryanan – Kinetic expts > large batch processing? Dynamics at low-q? (Paul's note: not sure what this means?)(Steve's note: My guess would be this is multi-frame data. 5.0.5 expanded the loader to read CanSAS files with multiple SASdata blocks but I've not explored what SasView then does with those, presumably it populates the Data Explorer with those multiple datasets which you can then batch fit?)
• Trevor – fibre diffraction tools SK to pick up with Trevor (Paul's note: was this resolved? is this Corfunc?)(Steve's note: No, not Corfunc. I talked to Trevor about this at Code Camp IX (2019). It was more about the 2D image pre-processing - like roving window background corrections - and then 2D image interpretation tools for indexing, etc. Now on the one hand we could say this is raw data operations and not for SasView, on the other hand we could say some of this is advanced slicer functionality? But, my feeling was that the holistic way to deal with this would be in Mantid, and in fact there is probably a lot of what Trevor wants already in Mantid). FINAL NOTE: agreed nothing to add a this point.
• Steve - how much is FIT2D still used? Answer: a lot (not necessarily at ESRF)! Issue is whether SasView should try to support FIT2D formats, or FIT2D support, say, NXcanSAS? Jerome wants more info. What is the action here? SK send Jerome links to NXcanSAS? (Paul's note: was this resolved?)(Steve's note: I did have discussions with Jerome & Narayanan during/after Code Camp IX (2019). Jerome said he would look to implement the CanSAS formats in his FabIO library, but looking at the docs for that it's not clear if he has. Fit2D has not been updated since 2016. I think reading 1D FIT2D ASCII into SasView would probably work - though I'm not sure we have an example to prove that! - and the Image Converter will read TIFF. Most other FIT2D formats are raw formats. So in the absence of strong community requests to do more, I don't see this as a Roadmap item, and I rather think there is more of an onus on ESRF to join the modern era) FINAL NOTE: agreed nothing to add a this point.
• Adrian – relating orientation axes to experiments > more advanced training/tools (Paul's note: Do we need to work on PK tool? is that sufficient?)(Steve's note: Well, for a start we need to put the orientation viewer into 5.x... - Paul's note: done added to list)

[wpotrzebowski]

Strike out means item has been moved to appropriate issue (#7 or #6 )

Few additional items that may or may not be included:

Generic Scattering Calculator:

• Adding hydration layer -> too specific for RoadMap as an item but handled at the higher level
• Adding structure factor (beta approximation) to a particle (Yun's student project) -> already on the RoadMap
• Converting to perspective (instead of tools)? -> too specific for road map. TBD when it will be handled

Infrastructure/Packages:

• Making sascalc and sasview available on pypi -> can be added to existing item
• Continue supporting conda package -> for discussion with developers and external stakeholders
• Repository/Marketplace for Notebooks -> for discussion

Others:

• Adding full history stack (Redo/Undo functionality)
• Integration with external databases and metadata catalogs (e.g. SASBDB)

[butlerpd]

Everything moved to appropriate list. Can no close this issue