Items remaining to be incorporated into the next iteration of the roadmap

[butlerpd]

• Refactor the way plotting is tied to data and fitting etc
• Refactor reporting and exporting of perspective data
  • @smk78 says: as Lucas has started this, aim for 5.1?
• Refactoring of project/analysis Saving and Loading
• Slicer refactoring (including box sum, new slicers,2D peak fit for centroid etc)
• Improve magnetic SANS usability/utility
• upgrade model marketplace
• refactor handling of multiple data sets for various use cases/workflows in conjunction with NXcanSAS
• Add support for handling multiple scattering
• Extend available integration options such as adaptive, MC integration etc
• Implement custom workflows (such as: finding the contrast match point from a series of datasets; pulling out the nuclear vs magnetic scattering from several polarised 2D datasets or from 1D cuts parallel and perpendicular to the field, etc.)
• P/*S computation enhancements (adding locally monodisperse approx.; fix volume fraction issue, etc.)
• reimplement the orientation viewer to 5.x (a 5.1 task for feature parity)
• simplify dependency/requirements issues (minimalize number/ well supported/ single source for all requirements (for all platforms and release vs development etc)
• Add full history stack (Redo/Undo functionality)
• Add proper Qz support to SasView analysis
• Add tutorials (written an video) for scripting (replaces "quick start guide for")
• Add tutorials (written and video) for writing models (replaces "quick start guide for")
• add external database integration for final results deposition (e.g. SASBDB and/or retrieval of data from various database catalogues (e.g. scicat, SASBDB, etc.)

Need to edit the relevant lines in the roadmap to make clear they cove these as well

• Implement calculation of 2D scattering patterns from oriented systems (https://doi.org/10.1107/S1600576717005179)
  • This looks like it would be another builder/engine for the general scattering calculator like sponge and mcsim and the current oriented engine (which may be wrong). If so this should be part of that bullet in the roadmap rather than a separate bullet.
• Clean up resolution calculations (ensure each q has a basis of support in calculated q values, then submit that full vector to the calculator as one piece; the current method computes the end pieces separately from the middle)
  • while no doubt useful, as stated it seems much too granular for the roadmap?
  • @smk78 says: this could be part of a larger resolution task with implementing user-defined resolution functions? Also need to fix the problem with smearing not being saved/loaded in projects!
• include the beta approximation calculation as example of other algorithms to add to the generic scattering calculator.
• Make sure it is clear that part of "creating a separate package (sascalc, data loader etc) they package should go on pypi
• Make clear that Dataloader package separation includes working with NXcanSAS to better utilize multiple entries for kinetics vs multiple detectors on single instrument vs multiple configurations vs 2 to 4 magnetic cross-sections etc. (alternatively could make this a separate entry?)
• include custom cost functions (other than Xi^2) and/or intensity weighting (beyond sigma I or sqrt(I)) as potential ways to address multimodal analysis?

Questions

• Is starting to implement PRISM really a task for 5.1, or later?
  • @butlerpd says: only because in talking to Tyler he says it (it = pyprism) is technically ready now to integrate (he thinks) and he is considering coming to the October code camp. Otherwise no, it would not make sense.

[butlerpd]

OWNERSHIP of Talking Stick

• Wojciech
• Steve
• Paul