A mixture M consists of an integer m≥1, real numbers aj,wj indexed by j∈{0,…,m−1}, and a real number λ, with 0<aj<1, wj>0 for every j, and λ>0; no condition ∑jwj=1 is imposed. For real c,t, define Qc(t)=((1+t)−c+(1−t)−c)/2, where powers are the code's total real-power operation. For real t, let CM(t)=∑j=0m−1wjaj/(1−ajt2) and DM(t)=λ/(1−t2); for real c,p,t, let eM,c,p(t)=(pQc(t)−CM(t))/(pQc(t)+DM(t)), with total real division. Let EM(c,p)={∣eM,c,p(t)∣:0≤t<1} and rM(c,p)=sSupREM(c,p), the total real supremum operation, which uses its library default on an empty or unbounded-above set. A pair (c,p)∈R2 is admissible exactly when 0≤c≤2 and p>0. For real c,t, define hc(t)=∫0tQc(s)ds, the oriented interval integral with respect to Lebesgue measure. For every real t, write U(t)=(log(1+t)−log(1−t))/2, using the total real logarithm of the code. Define SM(t)=∑j=0m−1wjajU(ajt) for real t. For real c,p,ℓ, a natural number N (possibly zero), and a sequence x:N→R, an N-step trajectory means that −1<xi<1 for every natural number i≤N, and that phc(xi+1)+λU(xi+1)=phc(xi)−SM(xi)−ℓ for every natural number i<N; no condition is placed on xi for i>N. For every real t, write H(t)=((1+t)log(1+t)+(1−t)log(1−t))/2, using the total real logarithm of the code. Define FM(t)=∑j=0m−1wjH(ajt) for real t. For real ℓ,t, let fM,ℓ(t)=FM(t)+λH(t)+ℓt and let GM,ℓ(t) be the actual total derivative operator deriv(fM,ℓ)(t) used in the code. For real c,p and natural N, let PM(c,p,N) be the set of all real numbers ∣GM,ℓ(xN)∣2/∣GM,ℓ(x0)∣2 obtained from some real ℓ and some N-step trajectory x with GM,ℓ(x0)=0; let ΩM(c,p,N)=sSupRPM(c,p,N), using the total real supremum operation even when the set is empty or unbounded above. A pair z=(c,p) is an optimum for M exactly when it is admissible and rM(c,p)≤rM(d,q) for every admissible pair (d,q). Let B0(M)=∑j=0m−1wjaj, B1(M)=∑j=0m−1wjaj/(1−aj), amax(M)=sSupR{aj:0≤j<m}, τ(a)=(1+8a−1)/2 for real a, and β(M)=B0(M)/(B0(M)+2λ); the supremum defining amax(M) is over a nonempty finite set. Assume the following assertions with their full universal quantification: For every mixture M, every real c,p with (c,p) admissible, and every natural number N≥1, the real number rM(c,p)2N is a least upper bound of PM(c,p,N): every member of that set is at most rM(c,p)2N, and rM(c,p)2N is at most every real upper bound of that set; and For every mixture M and every admissible real pair (c,p), rM(c,p) is a least upper bound of EM(c,p): every ∣eM,c,p(t)∣ with 0≤t<1 is at most rM(c,p), and rM(c,p) is at most every real upper bound of all these values; and For every mixture M, there exists exactly one real pair z that is an optimum for M, and every such optimum z=(c,p) satisfies 0<c<τ(amax(M))<1, B0(M)/2<p<B1(M), and 0<rM(c,p)<β(M)<1. Under these assumptions, the theorem concludes the following assertion with its full universal quantification: For every mixture M, every natural number N≥1, and every real pair z=(c,p) that is an optimum for M, ΩM(c,p,N)≤ΩM(d,q,N) for every admissible pair (d,q), and every admissible pair (d,q) with ΩM(d,q,N)=ΩM(c,p,N) equals (c,p).
Preamble
import Definitions.Def_PGDMap20261031_Entropy
open PGDMap20261031
Jointly Optimal Bregman Geometry and Stepsize for Positive Entropy Mixtures, manuscript dated 2026-10-30, Theorem 1(4), p. 2; proof pp. 4-5; local source: 03_Jointly_Optimal_Bregman_Map_and_Stepsize.pdf