Example. Observables on a finite configuration set [fdip-0022]
Example. Observables on a finite configuration set [fdip-0022]
Let \(V\) be a finite vertex set and \(S\) a finite set of possible values. A configuration is a function \(\sigma :V\to S\); the configuration set is \(\Omega =S^V\). The set \(\Omega \) is distinct from the vertex set \(V\): an observable takes a whole configuration as its input. For a commutative unital ring \(R\), set \[A_\Omega =\operatorname {Map}(\Omega ,R).\] This is the pointwise function algebra: \((f+g)(\sigma )=f(\sigma )+g(\sigma )\), \((fg)(\sigma )=f(\sigma )g(\sigma )\) and \((rf)(\sigma )=rf(\sigma )\), with unit the constant function \(1_R\).
For each \(\sigma \in \Omega \), evaluation \(\operatorname {ev}_\sigma (f)=f(\sigma )\) is a unital \(R\)-algebra map and a normalized linear functional. The scalar map sends \(r\) to the constant function \(r\). The indicator \(\delta _\sigma \) is \(1_R\) at \(\sigma \) and \(0_R\) elsewhere. These are the same indicators used in counting duality for finite functions, now indexed by configurations.
If \(V=\varnothing \), then \(S^V\) contains the unique empty function, even when \(S=\varnothing \). If \(S=\varnothing \) and \(V\ne \varnothing \), then \(\Omega =\varnothing \); its function algebra is the zero algebra. There is then no evaluation at a configuration.
The underlying unital algebra convention is
[stacks2017stacks, Section 22.2], Tag 09JF.
Mathlib's Pi.algebra, Pi.evalAlgHom and
Pi.constAlgHom provide the pointwise algebra, evaluation and
constant algebra maps. The configuration interpretation and the empty-set
cases above are direct consequences of these definitions.