Ellipsoid?

[yakir12]

Is the sphere supposed to be a subset of an ellipsoid? If and when we allow for transformations like mentioned in #11 (maybe with CoordinateTransformations.jl) then we'll need to connect a Ellipse to a affine transformed sphere. I might be totally off about what this package should include...

[SimonDanisch]

The scope of this package is supposed to be pretty broad, but in reality it mainly contains things I needed at some point ;) Ellipsoid definitely fits the scope, but I personally don't have any use for it! Feel free to add it :)

[yakir12]

OK, I'll consider. Right now (once I distro-hop to manjaro to get makie to work on my laptop, I'm him) I'll make do with affine-transforming a sphere to an ellipsoid.

[yakir12]

This is partly about PR #202 but mostly about the generalization of spheres to ellipsoids (of n dimensions). While this PR (currently incomplete) offers a new type, HyperEllipse, that is orthogonal to HyperSphere, one could imagine how spheres would be some specific instance of an ellipsoid. There are many ways to implement something like that. I was just about mention how currently a square is implemented just as a Rectangle whose sized are equal (and likewise a circle could be an ellipse who's sides are equal), but then I found out that there is a Cube type (but no Square)! That seems a bit inconsistent.

In any case, is there a need or interest in defining an Ellipse with specialized methods for circles (or some other type/trait hierarchy)? Or should we keep the status quo, and (once finished) merge this orthogonal implementation of an ellipse, #202 (that bares no relationship to a circle and vice versa)?