Path-connectedness of the cube and its boundary #
Mathlib's higher homotopy groups π_ n X x are built from generalized loops Ω^ N X x,
continuous maps I^N → X sending the cube boundary Cube.boundary N
({y | ∃ i, y i = 0 ∨ y i = 1}) to the base point. Reasoning about π_ n for n ≥ 2
needs to know how the cube and, crucially, its boundary are connected: the boundary of an
n-cube is the topological sphere S^{n-1}, which is connected precisely when n ≥ 2.
Mathlib records the cube boundary set but proves nothing about its connectivity.
This file supplies that missing input:
- the whole cube
I^Nis path connected (TauCeti.isPathConnected_cube); - its boundary is path connected as soon as the index type has at least two elements
(
TauCeti.isPathConnected_cubeBoundary).
The resulting path-connectedness declarations expose JoinedIn witnesses through their
.joinedIn methods, so callers can use the generic connectedness API directly.
The paths are elementary: a coordinate is dragged to 0 along the straight line
pathTowardZero, and any boundary point is joined to the corner 0 in two phases that each
keep one coordinate pinned at an endpoint, so the whole journey stays inside the boundary.
The two-element hypothesis is exactly what lets the second phase pin a different coordinate
at 0 while releasing the first.
Main declarations #
TauCeti.pathTowardZero: the straight-line path inIfromato0.TauCeti.isPathConnected_cube:I^Nis path connected.TauCeti.zero_mem_cubeBoundary: the corner0lies on the boundary.TauCeti.isPathConnected_cubeBoundary: for[Nontrivial N],Cube.boundary Nis path connected.
The straight-line path in the unit interval I from a to 0, given by t ↦ a * σ t
where σ is the interval symmetry t ↦ 1 - t.
Equations
- TauCeti.pathTowardZero a = { toFun := fun (t : ↑unitInterval) => a * unitInterval.symm t, continuous_toFun := ⋯, source' := ⋯, target' := ⋯ }
Instances For
The cube I^N is path connected: every point is joined to the corner 0 by the pointwise
product of the coordinate paths pathTowardZero.
The corner 0 lies on the boundary of the cube (any coordinate is 0).
The boundary of the cube I^N is path connected once the index type has at least two
elements. Any boundary point y, extreme in some coordinate i₀, is joined to the corner 0
in two phases: first collapse every other coordinate to 0 (keeping i₀ extreme), then
collapse the i₀-coordinate (keeping a different coordinate j₀ at 0). The two-element
hypothesis provides the index j₀ ≠ i₀.