Revisit the initial requirements

[yxqd]

The original code by Max Kresch works but it is showing some age and users complain about some missing features and ask for a better user interface. Hence a meeting with requirements for a new implementation of the getdos code.

from https://github.com/sns-chops/multiphonon/blob/2c0cb427061af44e26a279c8008839ee9092265d/docs/initial_requirements.jpg

These requirements were gathered at the start of this project and some of them were implemented while others are not. This ticket revisits the list and check off those items that are done, and new tickets will be created for items still pending.

The list include items intended to fix problems in the original code and items for desired new features

Fix problems

• Remove pkl files: pkl files are not used in the new code. hdf5 histogram files are used instead.
• Hide divide by zero warning: See hide divide by zero warning in dos2sqe calculation #33
• Remove the requirement that E=0 has to be center of an energy bin: Remove the requirement that E=0 has to be center of an energy bin #39

Desired features

• Use elastic line to estimate BG correction: See Subtract the elastic line from the MT-corrected SQE before converting SQE to DOS #34
• Mantid support: See Support Jen's mantid algorithm #16
• Work with multiple Eis: See sqe2dos: support S(Q,E) that only measures a portion of the DOS #30
• Save intermediate results: already implemented
• Handle slight/trivial mismatch of MT and sample SQE arrays: this is supported as a interpolation method is implemented. A user can interpolate the one SQE histogram to the other. See https://github.com/sns-chops/multiphonon/blob/2c0cb427061af44e26a279c8008839ee9092265d/multiphonon/sqe/__init__.py#L8
• Include SQE data in full Q range in calculation: this is already implemented.
• Able to compare to DFT result: this is outside the scope of this software I think. If a DFT calculation already exists, and the user wants to use it as the initial guess for DOS calculation, it can be done with the initdos argument: https://github.com/sns-chops/multiphonon/blob/2c0cb427061af44e26a279c8008839ee9092265d/multiphonon/getdos.py#L12
• Options to not do multiphonon correction: this can be achieved by using singlephonon_sqe2dos
• Reuse analysis settings in multiple analysis runs (for example, temperature series measurements): see support batch mode #35

[yxqd]

The tickets remained from this initial requirements after 0.1.1 release are

• Support Jen's mantid algorithm #16
• Subtract the elastic line from the MT-corrected SQE before converting SQE to DOS #34