Chain complexes, boundary maps, and the key identity ∂² = 0
Take the boundary of a triangle, then the boundary of that, and everything cancels. That one fact, ∂² = 0, is the engine that makes homology work.
Click any vertex, edge, or triangle to see its boundary with alternating signs. Then click “Apply ∂ twice” and watch the terms cancel in pairs.
The boundary operator ∂ maps each simplex to its boundary: a triangle maps to the sum of its three edges (with signs), an edge maps to the difference of its endpoints. The remarkable identity ∂² = 0— the boundary of a boundary is always zero — is the algebraic engine that makes homology possible.
A chain complex packages these boundary maps into a sequence of vector spaces connected by linear maps. Homology lives in the gap between cycles (things with zero boundary) and boundaries (things that are the boundary of something).
Key insight: The sign alternation in the boundary formula is not arbitrary — it is precisely what makes ∂² = 0 work. Each vertex appears in two boundary terms with opposite signs, so they cancel. This algebraic miracle is the foundation of homology theory.
Pick a space and switch between the views. Ker / Im shows which chains are cycles and which are boundaries; Homology shows what is left over.
A chain complex is a sequence of groups connected by boundary maps: … → C&sub2; → C&sub1; → C&sub0; → 0. Toggle to kernel/image view to see which elements are cycles (ker ∂) and which are boundaries (im ∂). Homology is the quotient: cycles modulo boundaries.
Key insight: Since ∂² = 0, every boundary is automatically a cycle (im ∂ ⊆ ker ∂). The homology group Hₖ = ker ∂ₖ / im ∂ₖ₊¹ measures how many cycles are not boundaries — these are the “holes.”
Pick a boundary matrix and step through the row and column operations until it is diagonal. Zero entries give free ℤ summands; watch for the 2 in RP², which gives torsion ℤ/2ℤ.
| e₀₁ | e₁₂ | e₀₂ | |
|---|---|---|---|
| v₀ | -1 | 0 | -1 |
| v₁ | 1 | -1 | 0 |
| v₂ | 0 | 1 | 1 |
The boundary operator can be written as a matrix. Reducing it to Smith normal form (the integer analogue of row echelon form) reveals the homology groups directly: zero diagonal entries correspond to free generators (ℤ), while nonzero entries reveal torsion (ℤ/nℤ).
Key insight: Smith normal form is the computational backbone of homology. The diagonal entry “2” in RP²'s boundary matrix means a cycle wraps around twice before bounding — giving the torsion group ℤ/2ℤ in H&sub1;(RP²).
Spot a mistake or have an idea?