Dependency on "Resolution" variable is restrictive for .dat file input

[PhilipOesterlePekrun]

@RShaw026
@NoraHagmeyer

Currently, a lot of the code is built around the "Resolution" variable even when using a .dat file for the height field, so this height field always needs to have dimensions (2^Resolution+1) × (2^Resolution+1) with some integer Resolution. But, it seems to me that the "Resolution" is only really relevant for the random midpoint algorithm to generate random topology. So, I think it would make sense to not restrict the height field in this way in general, and simply base it off N (N × N cells/elements/height-field nodes).

The only issue with this is that the shape factors are given discretely in terms of the Resolution. But, the shape factors aren't fundamentally tied to a resolution and could also be given as functions of N, if I understand correctly. As a first step, the shape factors could simply be interpolated.

Let me know your thoughts on this. I am also restructuring the overall input procedure a bit to make it more straightforward in preparation for Kokkos integration.

[mayrmt]

@RShaw026 @jacopenji @isteinbrecher Do you have an opinion on this topic?

[RShaw026]

Jacopo and I discussed this once. The idea is to interpolate the shape factors for the values of nodes other than 2^n+1.

[PhilipOesterlePekrun]

@RShaw026 That was what I was thinking, exactly. This shouldn't be too difficult, so I'll add this.

[isteinbrecher]

I also think that interpolation should work.

[PhilipOesterlePekrun]

Yep, it's implemented as part of #110

[mayrmt]

PR #110 has resolved the topic of this discussion.