Two

ToDo

Email Jean-Luc, Andrew, Max
- Thoughts about ARKs, NOIDs, NCAR Facilities and Instruments, Data Pipelines and stuff and things?

A Landing Page of Things and stuff.

Started the day with an example of what monte carlo can be used for to approximate Pi, and then do ray tracing for a detector.

McStas
- Monte Carlo Simulation of triple axis spectrometers
- Should be McStof (Time of Flight) ;) is a joke. isn’t going to happen.
Clinton Pile at Oak Ridge (one of the first production research reactors)
- Someone had the idea that we could cut some holes in the side and there we go. World’s first Neutron source.
- Nobel Prize for Neutron Scattering to these guys: Clifford Shull, Bertrand Brockhouse
  - For Establishing Neutron
- First place that neutron diffraction happened.
Work starded on McStas in 1997. Release in 1998. Peter started in 2002.
Bragg is a british physicist that established x-ray diffraction. (1894ish)
Kristian Nielsen is the original CS guy that taught the physicist
“Peter was explaining how McCode compiles a binary and runs it on HPC”
OpenACC
- NVidia compiler like “gcc but openmp”
- nvhpc is the only compiler that fully supports this.

Would be good to do it all on ESS-Compute
ORSO - Open Reflectometry Standards Organization - (Andrew is one of the co-chairs. They are trying to standardize across reflectometry)
- Agreed upon file format. (Just an ASCII file with a meta-data rich header)
- There is an organization called “Reflectivity”. (They had a merge request from one of the ORNL guys)
- “orsopy”
  - Python package that reads and writes ORSO specifications.
- They are moving towards HDF and then NeXuS - NeXus Interarnational Advisory Committee.
GenX - Reflectometry analysis software
There are about 50 different software packages for reflectometry analysis.
They are discussing writing the model for analysis as a seperate file.
Artefacts are whatever comes out of your analysis
“We’ve got raw data, and when you reduce the data, the reduced data becomes a child of the data-set (Derived DataSet)”
From the Derived DataSet you can have multiple artefacts (scripts, analysis methods, software, savable object from a gui taht describes the exact state of the system)

Solo project from Andrew M
Currently only used by two people.
With EasyReflectometry you can pick the “Engine”
- MotoFit (IgorPro)
- RefNX (Python Library)
- Refl1d (Mostly from NIST. NIST was quite keen on using EasyR to create models to put into refl1d)
- GenX

These things would have to be agreed upon and/or standardized as meta-data categories
- Initial state, potentially random seed
  - “There are some optimizations that are stochastic in nature and you need to know the initial seed”
- Model applied to the data
Neuten Raphsen
Reflectometry, Small Angle Neutron Scatterin (SANS), Quasi-Elastic Neutron Scattering (QENS)
Magnetic Inelastic Scattering (INS) might not adhere to this rough thing. Much harder to fit models in that space.

The plan is that everything will go through NICOS (authoritative data source)
- Then it can be added into the NeXus files.

25 Years ago they decided to come up with a nexus standard
5 years ago it was adopted.
Andrew “I think SASView is the gold standard for Neutron Analysis Software. In terms of how the community is structured. It’s a good thing”

There used to be a list on lightsources.org. Not sure now.