Composing relative CW complexes #
Let C be a CW complex relative to D, and let D in turn be a CW complex relative to E, in a
Hausdorff space. Then C is a CW complex relative to E (TauCeti.relCWComplexTrans), whose
n-cells are the n-cells of D together with the n-cells of C, with their characteristic
maps (TauCeti.relCWComplexTransCellEquiv). Taking E = ∅ and passing to an absolute CW complex
with Topology.RelCWComplex.toCWComplex, a complex built by attaching cells to a CW complex is a
CW complex.
One compatibility condition between the two structures is needed. The boundary of an n-cell of
C must lie in finitely many cells of dimension less than n, so the part of it in D must lie
in the skeleton skeletonLT D n. This is the only hypothesis. It holds, for instance, when all
cells of D have dimension less than all cells of C, the situation of a tower of cell
attachments of increasing dimension. Without it the cells of D and of C together need not
satisfy the closure-finiteness axiom: attaching a 1-cell to the closed 2-disk, with its usual
cells, along two interior points of the disk gives a CW complex relative to the disk, but the
boundary of the 1-cell lies in the open 2-cell.
Given the condition, the boundary of a cell of C is compact, so it meets only finitely many open
cells of D (TauCeti.finite_setOf_nonempty_inter_openCell_of_isCompact); those of them that
meet the boundary have dimension less than n. The weak topology of the composite combines the
weak topologies of the two structures, because every closed cell of D lies in D.
Main definitions #
TauCeti.relCWComplexTrans: the CW structure onCrelative toE.TauCeti.relCWComplexTransCellEquiv: itsn-cells are then-cells ofDand ofC.
References #
- A. Hatcher, Algebraic Topology, Chapter 0 for CW complexes built by attaching cells inductively, and the Appendix, Proposition A.1, for the compactness argument.
Composing relative CW complexes. If C is a CW complex relative to D, and D is a CW
complex relative to E, then C is a CW complex relative to E, as soon as the part in D of
the boundary of each n-cell of C lies in the skeleton skeletonLT D n. Its n-cells are the
n-cells of D together with those of C, with the same characteristic maps.
The structure is not an instance: the base of a relative CW complex is an outParam, so C
cannot carry one instance for each base.
Instances For
The n-cells of C relative to E are the n-cells of D together with the n-cells of
C.
Instances For
The characteristic map of a cell of D, as a cell of C relative to E, is its
characteristic map in D.
The characteristic map of a cell of C, as a cell of C relative to E, is its
characteristic map relative to D.
The open cells of C relative to E coming from D are the open cells of D.
The open cells of C relative to E coming from C are the open cells of C relative to
D.
The closed cells of C relative to E coming from D are the closed cells of D.
The closed cells of C relative to E coming from C are the closed cells of C relative
to D.
The boundaries of the cells of C relative to E coming from D are the boundaries of the
cells of D.
The boundaries of the cells of C relative to E coming from C are the boundaries of the
cells of C relative to D.