What is manipulations.py? and what can be deleted?

[lucas-wilkins]

I'm just looking at 'sascalc.dataloader.manipulations' - some of it doesn't seem finished, and its really hard to figure out what its actually for. It looks quite old.

Is this being actively worked on? Can I just delete things that are not used anywhere?

[butlerpd]

I believe that is some duplication of the main dataloader stuff in sasview so that folks using sasmodels don't need to import all of sasview (or eventually even just sascalc) in order to be able to use sasmodels? There is an idea to migrate the reader stuff to its own package as described in the sasdata repo which is a stub for such work as far as I can tell.

I'm guessing @krzywon or maybe @pkienzle would be the ones who could best answer the question on current status?

[smk78]

I'm guessing @krzywon or maybe @pkienzle would be the ones who could best answer the question on current status?

Both were on todays call and @lucas-wilkins and @krzywon are going to talk offline about this.

[butlerpd]

OOpss. I'm not sure why this is in the loader. This is where all the calculation for the slicers get done. Those are (aspirationally perhaps) a work in progress. There is a branch working on some fixes to slab slicers (though mostly the GUI side I think) that was again discussed at the magSANS workshop for other reasons. There are also I believe some issues to improve some of the calculations for the sector averaging (which currently does not support folding) and a number of the issues submitted at the ESS training workshop that would go here (all the rocking curve stuff basically).

That said I am not familiar with the code in the first half of that module.....

[lucas-wilkins]

Abandoned, but there is work in progress on related thing. Decide later.

[butlerpd]

Also look at moving to data-utils as part of moving data loader package.

[butlerpd]

Working on cleanup in preparation for Contributor Camp XII:

@lucas-wilkins and @krzywon I think we can close this ticket now? It seems totally obsolete though there may be some more specific issues related to the new manipulation.py but I presume those would best be addressed as separate issues?

[krzywon]

Much of the cleanup for this is in #47 started by @ehewins. This issue should probably be linked to that PR.