Terminology for the preconditionning

[sannant]

How to call the preconditionned objective ?

currently it is called mass_weighted, but it is actually preconditionning, and any person not familiar with Martin Müser's methods will not understand what the mass is about.

I would call it preconditioned instead of mass_weighted.

@pastewka , @sindhu2993 , opinions ?

[sindhu2993]

I would also agree with calling it 'preconditioned'. Provides a clearer sense of what is done inside.

[pastewka]

Call it preconditioned

[sannant]

A similar question also arises for how to call the variables in halfcomplex fourier space.

At the moment: force_k

My suggestion: force_halfcomplex or halfcomplex_force

Sometimes there is also k_float instead of k used, making clear it is the halfcomplex fourier field.

[pastewka]

The _k suffix should indicate that this is an array that has the wavevector as the index. Better would be "_kl" or alike (if it's 2D- but I guess the whole thing also works for 1D).

I would therefore also use a single letter suffix, i.e. use "_h" or "_q" for halfcomplex, but document it somewhere.

[sannant]

I think there is a risk that _q and _k are used as synonyms in part of the code.

Hence _h sounds as the reasonable choice.

[pastewka]

Sounds good!

[sindhu2993]

Okay so I update the force_k and force_k_float by force_h.

[sannant]

Ok feel free to deprioritize.

[sannant]

I still need to do this: rename the _k to _h when referring to functions working in halfcomplex space.