## Coordinate and Coordinate System for Data Frames

Need a container for describing bulk data. Looking at the XDR and related standards may be particularly helpful.

A coordinate system should be an abstract thing which is associated with a dataframe.

There are many considerations:
Explicit coordinates (e.g. every single element is treated as independent, so for NxM (y,x) frame you have N*M x coordinates and N*M y coordinates). This is effectively identical to the point cloud representation.
Implicit rectilinear coordinates (e.g. the coordinates are stored elsewhere and per-element coordinates derived based on indices of elements).

Global/absolute coordinate systems vs. object coordinate systems.

Different coordinate systems, conversion between representations. e.g. if I have a cartesian coordinate system, what if I want all of the elements converted into a polar coordinate system? What about cylindrical/spherical?

Element pitch - size of each element (e.g. pixel size)

Differential domain (e.g. is the pixel representing point-sampled data? Is it showing integrated densities? Is it representing gradients?)

Units - of coordinates (e.g. time (s) or time (us)) and of data elements (e.g. pressure, microvolts)

Domain (e.g. R, R^+, Complex numbers)

Gauge/metric (given two elements, how close are they in space. Separate but related, given two elements, how similar are they?)

Interpolation - To a specific set of coordinates, or to a specific number of elements (e.g. 256 sample points, interpolate to 350 over the same span).