A Closer Look on the Influence of Constraints Upon the Optimization of the Nonadditive Entropic Functional S q S_{q}

Leandro Lyra Braga Dognini, Constantino Tsallis · 2026

Read A Closer Look on the Influence of Constraints Upon the Optimization of the Nonadditive Entropic Functional S q S_{q} by Leandro Lyra Braga Dognini, Constantino Tsallis (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-08860.txt.

Introduction

Leandro Dognini and Constantino Tsallis rework how firm states come from a broad entropy max. The broad entropy has an index q in it. The mean energy rule can weigh states with a second index. That makes a clan of max tasks whose best spreads need care. The team sets terms for best spreads to be there, to be firm, and to be the sole ones. A claim lets some be worked out in closed form.

Two stock cases in the field grow clear. A new third case joins them. Just the stock cases give the famed q-exp shapes. A Clausius style heat scale and free energy come out too. The zeroth law of heat stands in this broad frame.

Catalog record

CREATORLeandro Lyra Braga Dognini, Constantino Tsallis
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE92 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.08860

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Closer Look on the Influence of Constraints Upon the Optimization of the Nonadditive Entropic Functional S q S_{q} Leandro Lyra Braga Dognini, Constantino Tsallis Abstract The thermal-equilibrium canonical distribution is currently obtained by maximizing the Boltzmann-Gibbs-von Neumann-Shannon entropy S B ​ G ​ ( p ) = k ​ ∑ i = 1 W p i ​ ln ⁡ 1 / p i S_{BG}(p)=k\sum^{W}_{i=1}p_{i}\ln 1/p_{i} constrained to ∑ i = 1 W p i = 1 \sum^{W}_{i=1}p_{i}=1 and ∑ i = 1 W p i ​ e i = U \sum^{W}_{i=1}p_{i}\,e_{i}=U , e 1 ≤ … ≤ e W e_{1}\leq\ldots\leq e_{W} being the energies of the W W possible states and U ∈ [ e 1 , e W ] U\in[e_{1},e_{W}] their mean value. We revisit a generalized version of this optimization problem grounded in the nonadditive entropy S q ​ ( p ) = k ⁡ ( ∑ i = 1 W p i q − 1 ) / ( 1 − q ) S_{q}(p)=k\,(\sum^{W}_{i=1}p_{i}^{q}-1)/(1-q) (frequently, though not necessarily, q ∈ ( 0 , 1 ) q\in(0,1) ; S 1 = S B ​ G S_{1}=S_{BG} ), and the constraint ∑ i = 1 W p i q ′ ​ e i / ∑ i = 1 W p i q ′ = U \sum^{W}_{i=1}p_{i}^{q^{\prime}}e_{i}/\sum^{W}_{i=1}p_{i}^{q^{\prime}}=U , q ′ > 0 q^{\prime}>0 . Sufficient conditions for existence, strict positivity, and uniqueness of solutions are derived, along with a theorem that enables their closed-form calculation. We apply these results to deepen the understanding of the two standard cases in the literature ( q ′ = 1 q^{\prime}=1 and q ′ = q q^{\prime}=q ), as well as of a new one ( q ′ = 2 − q q^{\prime}=2-q ). We prove that these standard cases are the only ones yielding optimizing probability distributions of q q -exponential form. Furthermore, we define an effective temperature T q , q ′ T_{q,q^{\prime}} through a Clausius-like relation 1 / T q , q ′ = ∂ S q / ∂ U 1/T_{q,q^{\prime}}=\partial S_{q}/\partial U and derive a Helmholtz-like energy F q , q ′ = U − T q , q ′ ​ S q F_{q,q^{\prime}}=U-T_{q,q^{\prime}}S_{q} , with the former grounding the validity of the 0 t ​ h 0^{th} Principle of Thermodynamics within this generalized statistical mechanics. Finally, we show that the case with a linear constraint (i.e., q ′ = 1 q^{\prime}=1 ) with q ∈ ( 0 , 1 ) q\in(0,1) (i) preserves the Third Law of Thermodynamics; (ii) can be used to model classical many-body Hamiltonian systems with arbitrarily-ranged interactions; and (iii) resembles features of low-dimensional nonlinear dynamical systems at the edge of chaos. 1. Introduction Standard statistical mechanics (herein referred to as BG statistical mechanics [ 1 , 2 , 3 ] ), grounded on the Boltzmann-Gibbs-von Neumann-Shannon additive entropic functional S_{BG} defined below, constitutes one of the pillars of contemporary theoretical physics. This superb theory satisfactorily handles, as is well known, a plethora of physical systems. However, strictly speaking, it fails when space and/or time long-range correlations are relevant, which is currently the case for wide classes of so-called complex systems, whether natural, technological, or social ones. By failure, we naturally mean that it does not quantitatively reproduce the corresponding experimental, observational, or computational data. A generalization of the BG theory was proposed in 1988 [ 4 ] by introducing nonadditive entropic functionals that generalize the BG one. This proposal, currently referred to in the literature as nonextensive statistical mechanics , has been submitted to careful scrutiny by the physics community for decades: details are available in [ 5 , 6 ] and in the Bibliography . The term “nonextensive” refers to the fact that long-range-interacting Hamiltonians constitute paradigmatic examples of such systems; in such cases, the total energy can be superextensive in the thermodynamic sense, in contrast to usual BG systems, where the total energy is extensive. The analytical formulation of this generalized theory is basically presented in [ 4 , 7 , 8 ] . However, a variety of details concerning its mathematical status have not been completely addressed in a unified approach. The goal of the present paper is to revisit the entire theoretical formalism by closely examining the influence of the constraints under which the optimization of the entropic functional is performed. The thermal-equilibrium canonical distribution is currently obtained through the maximization of the Boltzmann-Gibbs-von Neumann-Shannon entropic functional [ 1 , 2 , 3 ] under two constraints: one related to probability normalization and the other to the mean energy of the system. This optimization problem is written as: \displaystyle\max \displaystyle S_{BG}(p) (1) s.t. \displaystyle p_{i}\geq 0,1\leq i\leq W \displaystyle\sum^{W}_{i=1}p_{i}=1 \displaystyle\sum^{W}_{i=1}p_{i}e_{i}=U, with the entropic functional S_{BG}:\mathbb{R}^{W}_{+}\rightarrow\mathbb{R} 1 1 1 We write \mathbb{R}^{W}_{+}\equiv\{x\in\mathbb{R}^{W}\mid x_{i}\geq 0,1\leq i\leq W\} and \mathbb{R}^{W}_{++}\equiv\{x\in\mathbb{R}^{W}\mid x_{i}>0,1\leq i\leq W\} . given by \displaystyle S_{BG}(p)=k\sum^{W}_{i=1}p_{i}\ln\frac{1}{p_{i}}, k>0 being a positive constant chosen once and for all (usually k=k_{B} , the Boltzmann constant, in physics, and k=1 in computational sciences); W\geq 2 the number of possible states of the system; e_{1}\leq\ldots\leq e_{W} the energy levels of each state (all such values form the energy spectrum ); and e_{1}\leq U\leq e_{W} the mean or internal energy 2 2 2 Notice that, for two independent systems A and B (i.e., p_{ij}^{A+B}=p_{i}^{A}p_{j}^{B} , 1\leq i\leq W_{A} , 1\leq j\leq W_{B} ) with energy spectra of the joint system given by e^{A+B}_{ij}=e^{A}_{i}+e^{B}_{j} , the internal energies U^{A}=\sum_{i}p_{i}^{A}e_{i}^{A} , U^{B}=\sum_{j}p_{j}^{B}e_{j}^{B} and U^{A+B}=\sum_{i,j}p_{ij}^{A+B}e_{ij}^{A+B} are additive as well (i.e., U^{A+B}=U^{A}+U^{B} ), so that microscopic independence and additivity lead towards macroscopic additivity. . We recall that the degeneracy function is given by g(e_{i})=\#\{1\leq j\leq W\mid e_{j}=e_{i}\} , 1\leq i\leq W . Also, we straightforwardly verify that the entropic functional S_{BG} is additive , meaning that, for two independent systems A and B (i.e., p_{ij}^{A+B}=p_{i}^{A}p_{j}^{B} , 1\leq i\leq W_{A} , 1\leq j\leq W_{B} ), the entropy of the joint system A+B satisfies S_{BG}(p^{A+B})=S_{BG}(p^{A})+S_{BG}(p^{B})\,. Nonadditive entropies are a natural way to build generalizations of this optimization problem in order to tackle physical phenomena that are not well addressed by the Boltzmann-Gibbs paradigm. For instance, a class of optimization problems that generalizes ( 1 ) through the nonadditive entropic functional S_{q}:\mathbb{R}^{W}_{+}\rightarrow\mathbb{R} , q\in\mathbb{R} , \displaystyle S_{q}(p)=k\sum^{W}_{i=1}p_{i}\ln_{q}\frac{1}{p_{i}}=k\biggr(\frac{\sum^{W}_{i=1}p^{q}_{i}-1}{1-q}\biggr), for p\in\mathbb{R}^{W}_{+} , is given by \displaystyle\max \displaystyle S_{q}(p) (2) s.t. \displaystyle p_{i}\geq 0,1\leq i\leq W \displaystyle\sum^{W}_{i=1}p_{i}=1 \displaystyle\dfrac{\sum^{W}_{i=1}p_{i}^{q^{\prime}}e_{i}}{\sum^{W}_{i=1}p_{i}^{q^{\prime\prime}}}=U, with q,q^{\prime},q^{\prime\prime}>0 3 3 3 The case q<0 can, in fact, be similarly discussed; however, we will not address it in the present discussion. Let us only remind that S_{q} is to be maximized for q>0 , as in ( 2 ), and minimized for q<0 . . Clearly, S_{1}=S_{BG} and we remind that the q -logarithm \ln_{q}:(0,+\infty)\rightarrow\mathbb{R} , q\in\mathbb{R} , is defined by \displaystyle\ln_{q}z=\frac{z^{1-q}-1}{1-q}, for z>0 , hence \ln_{q}xy=\ln_{q}x+\ln_{q}y+(1-q)\ln_{q}x\ln_{q}y , for x,y>0 . We consistently verify that the entropic functional S_{q} , q\neq 1 , is nonadditive , meaning that, for two independent systems A and B , S_{q}(p^{A+B})=S_{q}(p^{A})+S_{q}(p^{B})+\frac{(1-q)}{k}S_{q}(p^{A})S_{q}(p^{B})\neq S_{q}(p^{A})+S_{q}(p^{B})\,. Notice that q=q^{\prime}=q^{\prime\prime}=1 in ( 2 ) recovers ( 1 ) and several other cases have been studied in the literature (see [ 9 , 10 ] for a comprehensive discussion on the solutions of these optimization problems). For instance, the case (q,q^{\prime},q^{\prime\prime})=(q,q,1) , q>0 , along with its

Read the complete text