ScalingStacks

Proof. [02N2]

Original official author HTML, exact retained edition. Historical TeX conversion verdicts remain unchanged. Cited-edition alignment and mathematical self-containment are not assessed.

Complete original source context · Original author HTML

Proof.

We start by proving (1). By the properties of uniform convergence, it is clear that any element of 𝒫¯​(Δ,Nℝ){\overline{\mathscr{P}}}(\Delta,N_{\mathbb{R}}) is concave and continuous. Conversely, a continuous function ff on Δ\Delta is uniformly continuous because Δ\Delta is compact. Therefore, given ε>0\varepsilon>0 there is a δ>0\delta>0 such that |f⁡(u)−f⁡(v)|<ε|f(u)-f(v)|<\varepsilon for all u,v∈Δu,v\in\Delta such that ‖u−v‖<δ\|u-v\|<\delta. By compactness, we can find a triangulation Δ=⋃iΔi\Delta=\bigcup_{i}\Delta_{i} with diam⁡(Δi)<δ\operatorname{diam}(\Delta_{i})<\delta. Let {bj}j\{b_{j}\}_{j} be the vertices of this triangulation and consider the function g∈𝒫⁡(Δ,Nℝ)g\in\mathscr{P}(\Delta,N_{\mathbb{R}}) defined as

g(u)=sup{∑j=1lλjf(bj)|λj≥0,∑jλj=1∑jλjaj=x}.g(u)=\sup\bigg\{\sum_{j=1}^{l}\lambda_{j}f(b_{j})\bigg|\ \lambda_{j}\geq 0,\sum_{j}\lambda_{j}=1\sum_{j}\lambda_{j}a_{j}=x\bigg\}.

For u∈Δu\in\Delta, let bj0,…,bjnb_{j_{0}},\dots,b_{j_{n}} denote the vertices of an element of the triangulation containing uu. We write u=λj0​uj0+⋯+λjn​ujnu=\lambda_{j_{0}}u_{j_{0}}+\dots+\lambda_{j_{n}}u_{j_{n}} for some λji≥0\lambda_{j_{i}}\geq 0 and λj0+⋯+λjn=1\lambda_{j_{0}}+\dots+\lambda_{j_{n}}=1. By concavity, we have

f⁡(u)≥g⁡(u)≥∑k=0nλjk​f​(ujk)≥f⁡(u)−ε,f(u)\geq g(u)\geq\sum_{k=0}^{n}\lambda_{j_{k}}f(u_{j_{k}})\geq f(u)-\varepsilon,

which shows that any continuous function on Δ\Delta can be arbitrarily approximated by elements of 𝒫⁡(Δ,Nℝ)\mathscr{P}(\Delta,N_{\mathbb{R}}).

We now prove (2). Let f∈𝒫¯​(Nℝ,Δ)f\in{\overline{\mathscr{P}}}(N_{\mathbb{R}},\Delta). By definition, for each ε>0\varepsilon>0 we can find a function g∈𝒫⁡(Nℝ,Δ)g\in\mathscr{P}(N_{\mathbb{R}},\Delta) with sup|f−g|≤ε\sup|f-g|\leq\varepsilon. In particular, |f−g||f-g| is bounded. Furthermore, rec⁡(g)=ΨΔ\operatorname{rec}(g)=\Psi_{\Delta} and |g−rec⁡(g)||g-\operatorname{rec}(g)| is bounded because g∈𝒫⁡(Nℝ)g\in\mathscr{P}(N_{\mathbb{R}}). Hence dom⁡(f)=dom⁡(g)=Nℝ{\operatorname{dom}}(f)={\operatorname{dom}}(g)=N_{\mathbb{R}} and |f−ΨΔ||f-\Psi_{\Delta}| is bounded.

Conversely, let ff be a concave function such that dom⁡(f)=Nℝ{\operatorname{dom}}(f)=N_{\mathbb{R}} and |f−ΨΔ||f-\Psi_{\Delta}| is bounded. Then stab⁡(f)=stab⁡(ΨΔ)=Δ\operatorname{stab}(f)=\operatorname{stab}(\Psi_{\Delta})=\Delta and f∨f^{\vee} is a continuous concave function on Δ\Delta. Hence we can apply (1) to f∨f^{\vee} to obtain functions gi∈𝒫⁡(Δ,Nℝ)g_{i}\in\mathscr{P}(\Delta,N_{\mathbb{R}}) approaching f∨f^{\vee} uniformly. We conclude that the functions gi∨∈𝒫⁡(Nℝ,Δ)g_{i}^{\vee}\in\mathscr{P}(N_{\mathbb{R}},\Delta) approach ff uniformly and so f∈𝒫¯​(Nℝ,Δ)f\in{\overline{\mathscr{P}}}(N_{\mathbb{R}},\Delta). ∎

Original mathematics by the credited authors. Source-backed reader collection; mathematical self-containment is not assessed.