Étale groupoids arising from products of shifts of finite type (Q323683): Difference between revisions

An étale groupoid \({\mathcal G}\) is defined as a second countable locally compact Hausdorff groupoid such that the range map is a local homeomorphism. The author formulates two conjectures (denoted HK and AH) about essentially principal (the interior of the isotropy bundle of \({\mathcal G}\) is \({\mathcal G}^{(0)}\)) minimal (every \({\mathcal G}\)-orbit is dense in \({\mathcal G}^{(0)}\)) étale groupoids \({\mathcal G}\) whose unit space \({\mathcal G}^{(0)}\) is a Cantor set and verifies these conjectures in a number of cases. HK claims that for such a \({\mathcal G}\), the homology groups \(H_n({\mathcal G})\) agree with the \(K\)-groups of the \(C^*\)-algebra \(C_r^*({\mathcal G})\): \(\bigoplus_{i=0}^{\infty} H_{2i} ({\mathcal G}) \cong K_0(C_r^*({\mathcal G}))\) and \(\bigoplus_{i=0}^{\infty} H_{2i+1} ({\mathcal G}) \cong K_1(C_r^*({\mathcal G}))\). AH claims that for such a \({\mathcal G}\), the abelianization \([[{\mathcal G}]]_{ab}\) of \([[{\mathcal G}]] \) (the topological full group of \({\mathcal G}\)) has a close connection to the homology groups \(H_0({\mathcal G})\) and \(H_1({\mathcal G})\): there exists an exact sequence \(H_0({\mathcal G})\otimes {\mathbb Z}_2 \to [[{\mathcal G}]]_{ab} \to H_1({\mathcal G}) \to 0\). It is proved that AH holds, for example, for the following groupoids: 1) principal, minimal, almost finite groupoids; 2) minimal purely infinite groupoids \({\mathcal G}\) for which the group \([[{\mathcal G}]]_0\) (the kernel of certain map of \([[{\mathcal G}]]\)) is generated by transpositions. For the second half of the paper, define first SFT groupoids. Let \((\mathcal{V, E})\) be a finite directed graph and \(A=(A( \xi, \eta))_ { \xi, \eta \in \mathcal V}\) the adjacency matrix of \((\mathcal{V, E})\). It is assumed that \(A\) is irreducible and that it is not a permutation matrix. Let \(X_A= \{(x_k)_ {k\in \mathbb N}\in \mathcal{E}^{\mathbb N}| t(x_k)=i(x_{k+1}) \forall k\in \mathbb N\}\), where \(i(x)\) and \(t(x)\) denote the initial and terminal vertex of \(x \in \mathcal E\), respectively, and let \(\sigma_A : X_A \to X_A\) be a map defined by \( \sigma_A(x)_k=x_{k+1} \), \(k \in \mathbb N\). The set \({\mathcal G}_A=\{(x,n,y)\in X_A\times \mathbb N \times X_A | \exists k,l\in \mathbb N, n=k-l, \sigma_A^k(x)= \sigma_A^l(y) \}\), together with suitably defined multiplication and inverses, is called the \textit{SFT groupoid}. It turns out that finite products of SFT groupoids satisfy HK and AH conjectures. It is also proved that if \({\mathcal G}\) is a finite product of SFT groupoids, then \(0\to S_0\otimes {\mathbb Z}_2 \to [[{\mathcal G}]]_{ab} \to H_1({\mathcal G}) \to 0\) is exact, where \(S_0\subset H_0({\mathcal G})\) is a certain subgroup.