Theorem 21.27.2. Let be a positive compactly supported Radon measure on of mass . There is a finite convex function with bounded and .
Argument in the manuscript
Proof. Choose finitely supported positive measures of the same total mass, supported in a fixed compact set , and converging weakly to . Such measures are obtained by partitioning into Borel sets of diameter tending to zero and assigning each set’s mass to one of its points. Ignore cells of zero mass. Apply Theorem 21.27.1 to obtain dual functions on .
All slopes of these maxima lie in . For any Euclidean norms on the dual vector spaces, there is therefore one constant such that every is -Lipschitz on . Fix and add constants to arrange . The functions are uniformly bounded by . Arzelà–Ascoli gives a subsequence converging uniformly on to a convex continuous function .
Let and , using extensions off . The elementary Legendre estimate gives Thus the convergence is uniform even on all of , and is bounded. Local uniform convergence of finite convex functions implies weak convergence of their Alexandrov measures. Equivalently, one can use the continuous real Bedford–Taylor theorem together with the equality . This standard convex-measure continuity is the one external analytic input in the limiting step. It yields . ◻