View source for /GISANS Data Reduction and Analysis Upgrades Initiative

=== Overview === 
'''GISANS Data Reduction and Analysis Upgrades Initiative:'''<br />
An initiative on “GISANS data reduction and analysis upgrades” was launched at the GISAXS-25 meeting in Hamburg in November 2025. The objective of the initiative is to develop a generic, instrument-independent software framework for GISANS data reduction and analysis that can be used across neutron research facilities.

'''Approach:'''<br />
The proposal is to build upon the open-source software package BornAgain and to identify and define the key developments required for broader GISANS usability. Planned areas of improvement include:
* more flexible treatment of instrument resolution,
* improved background subtraction,
* clearer approaches to normalization,
* increased interoperability with existing analysis and reduction software tools.

'''Planned Work:'''<br /> 
A dedicated working group has to be established with two primary tasks:
1. Defining a roadmap and priority list for the required functionality.
2. Implementing these features in collaboration with the BornAgain core development team. Contributions may include code development, example scripts, and other shared resources.

'''Next Steps:'''<br />
The GISANS community is being invited to participate in the initiative. A first workshop is planned for the first half of 2026, with Institut Laue-Langevin (ILL) proposed as the host venue. The meeting is intended to be held on-site, with the option for remote attendance.

=== Meeting 2025-11-13 ===

'''Participants:'''<br />
Adrian, Apostolos, Annika, Dirk, Artur. <br />
BornAgain team: Joachim, Mikail, Ammar. <br />
SAGA Team: Tom, Tommy, Max, Milan

'''Date:'''<br />
13th November 2025

'''Content:''' <br />
1.	Introduction (Annika)<br />
2.	Current activites in BornAgain (BA)  (Joachim Wuttke)<br /> 
3.	“Quick” Roadmap discussion<br />
4.	Preperation of on-site meeting at ILL<br />
5.	“How-to” <br />
 
 
====Introduction (Annika)====
*	Goal: Getting an instrument-independent GISANS software
*	Question: (For using BornAgain as such) How to make it attractive for more users? How to implement together more features?
*	Answers: Most important to (i) define most necessary items, (ii) keep it simple, (iii) have a core developers team (the current BA-team), that (iv) gets help by python/C++-code contributions by the community – e.g. into a separate repository that will be reviewed by the core team
*	Next steps: 2-3 days in-person meeting. Current idea: at ILL. Contributors: from each neutron source (experimentalists + data scientists) + the GISAS core users.

====Current activites in BornAgain (BA)  (Joachim Wuttke)====
*	BA is an open-source software, everyone can access and contribute
*	Idea: Make it simpler, cleaner, more accessible, in the best case community-maintained to get a long-term manageable project
*	Nearly everything is possible from Python interface, but not from the GUI.
*	The GUI shall not be altered too much for the next coming years
*	Plans: First achieving simplification, afterwards working on extensions
*	Current biggest challenge: handling of coherent vs. incoherent scattering contributions
*	Next goals: Auto-generation of new geometric shapes

====Preperation of On-site meeting at ILL====
*	Aim at first half of 2026
*	Ask for (i) core-team dates, (ii) community dates

====“Quick” Roadmap discussion====

'''POINT A.+B. ToF-GISANS + Multiple scattering'''<br />
*	Possible in Python / via jupyter notebook in BA. Keep it there.
*	Computationally heavy: has to be inserted into BA via slicing it down to monochromatic slices with individual resolution functions
*	Needs to be done via ROI-definition
*	Complicated for users:
**	especially due to complicated connection to multiple scattering & incoherent scattering
**	different backgrounds for different lambda/incident angles which is additionally dominated by multiple scattering
*	Need to differentiate between: (i) incoherent multiple scattering, (ii) multiple scattering for which Born Approximation breaks down
'''POINT C. Background handling'''<br />
*	A challenge for weakly scattering systems
*	Connected to points (A) + (B). 
*	Challenging case of angular-dependent “incoherent multiple scattering” in many soft matter systems 
*	The generic question “how do we subtract bkg?” becomes relevant as well
*	Again, computationally heavy/long. Possible solution: ROI-definition
*	Existing user scripts that have proven to work, can be implemented into BA

'''POINT D. Normalization''' <br />
*	In SAS, we put units of cm-1 (or sr-1). 
*	In reflectivity, data are normally reported as ratio. 
*	In GISAS, not clear consensus as to how data are reported
 
'''POINT E. Using SANS softwares''' <br />
* For data analysis: 
**	Have the possibility to import missing form factors from e.g. SASView, SASfit?
**	need a plug-in feature that is compatible with existing other software, e.g. via FabIO (Can it now read all detector formats into BornAgain?)
* For data reduction: 
**	Needs to be compatible to existing software like GRASP or MANTID
**  Key: well-defined .nxs (canSAS) standard 

'''POINT F / Instrumentation & Q/lambda resolution (different detector arrays, or, sample geometries-how is data reduced)'''<br />
*	Instrument-dependent aspects (more than 1 detector arrays with potentially useful and unexploited information, sample geometry, detector-background, polarization efficiency, error-bars) have to be reduced at the instrument as usual
*	Calculation of resolution function in proper coordinate system --> can be tricky for non-experts.
*	Important for implementation in BornAgain: Q/lambda resolution functions
**	Can be implemented fast (via “Issue Tracker” of BornAgain?)
**	Best to start from working/existing models
*	Polarization-dependent 4-channel calculation could be a pre-defined python script to make it easier for beginners
*	GISAXS community might be interested in joining the discussion on resolution functions
 
'''POINT G. Need a good documentation'''<br />
 
*	To be contributed via “requests”/”patches” to the documentation sources.
 
'''POINT H. Integration of McStas with BornAgain'''<br />
*	Feedback-loops available (from Artur) between McStas and BornAgain, but no fitting possible
  
'''POINT I. Get a common data format for GISANS data reduction:'''<br />
*	Making use of canSAS and ORSO standards
*	Need to extend both the canSAS/ORSO formats towards GISAS


====“How-to”====
*	Keep BA-group as core developers group. Contributions by the community on most urgent items
*	have a section on BornAgain source/repository (called contributions). People can put contributions there, not mixing with main core. This does not interferes with complicated C++ core of BornAgain, but would benefit from Python scripts.
*	Distribution of notes / goals / etc. + Conversation via cansas
**  Need to re-populate and extend the cansas wiki for GISANS
*	Occasional in-person workshops, first one at ILL first half of 2026 (aim at max ca. 40 people  can each facility cover the travel costs of the participating individuals?)
** Goal: Extending the roadmap + defining a priority list
** If possible, connection to another event would be appreciated – otherwise we organize it independently
** Partial financing by SAGA team possible for ca. 10 people
** Making online attendance possible
** If possible, including an overview of current status from BornAgain to all participants