`salvus.mesh.layered_meshing.utils.remap_xarray_to_unit_domain`

Remap a DataArray or Dataset slab onto the unit square (2D) or unit cube (3D).

Functions

`remap_xarray_to_unit_domain()`

def remap_xarray_to_unit_domain(
    da: _XrTypes, top: xr.DataArray | float, bottom: xr.DataArray | float
) -> _XrTypes: ...

Remap a slab of da between bottom and top onto the unit square/cube.

Per horizontal column the trailing axis is rescaled to t ∈ [0, 1] with bottom(x, y) → 0 and top(x, y) → 1. At t = 0 and t = 1 the output value equals da linearly interpolated along the trailing axis at the bounding surface — the “pin value” at the cut.

The output uses the same dim names as da with coordinate values replaced by np.linspace(0, 1, N_d); horizontal coordinates are re-labeled uniformly and so the input’s horizontal spacing is not preserved in the output coordinate values. The unit-space resolution along the trailing axis equals the densest column’s count of da grid points within [bottom, top], with a floor of 2; horizontal dims keep their input resolution. Pass top/bottom as a scalar for a flat cut (e.g. a grid-aligned boundary), or as a DataArray on the horizontal coords of da for a tilted/curved cut.

Parameters

da _XrTypes — The DataArray or Dataset to remap. Horizontal coordinates are the leading dimensions; the vertical coordinate is the trailing dimension and must be monotonically increasing.
top xr.DataArray | float — The upper bounding surface, as a scalar (flat cut) or as a DataArray on the horizontal coords of da (tilted cut).
bottom xr.DataArray | float — The lower bounding surface, with the same conventions as top.

Returns _XrTypes — The remapped DataArray or Dataset, on the same dimensions as da but with coordinate values uniformly spanning [0, 1].