Jump to content

q-Gaussian distribution

From Wikipedia, the free encyclopedia
q-Gaussian
Probability density function
Probability density plots of q-Gaussian distributions
Parameters q<3 shape (real)
β>0 (real)
Support x∈(−∞;+∞) for 1≤q<3
x∈[±1β(1−q)] for q<1
PDF βCqeq(−βx2)
CDF see text
Mean 0 for q<2, otherwise undefined
Median 0
Mode 0
Variance 1β(5−3q) for q<53
∞ for 53≤q<2
Undefined for 2≤q<3
Skewness 0 for q<32
Excess kurtosis 6q−17−5q for q<75

The q-Gaussian is a probability distribution arising from the maximization of the Tsallis entropy under appropriate constraints. It is one example of a Tsallis distribution. The q-Gaussian is a generalization of the Gaussian in the same way that Tsallis entropy is a generalization of standard Boltzmann–Gibbs entropy or Shannon entropy.[1] The normal distribution is recovered as q → 1.

The q-Gaussian has been applied to problems in the fields of statistical mechanics, geology, anatomy, astronomy, economics, finance, and machine learning.[citation needed] The distribution is often favored for its heavy tails in comparison to the Gaussian for 1 < q < 3. For q<1 the q-Gaussian distribution is the PDF of a bounded random variable. This makes in biology and other domains[2] the q-Gaussian distribution more suitable than Gaussian distribution to model the effect of external stochasticity. A generalized q-analog of the classical central limit theorem[3] was proposed in 2008, in which the independence constraint for the i.i.d. variables is relaxed to an extent defined by the q parameter, with independence being recovered as q → 1. However, a proof of such a theorem is still lacking.[4]

In the heavy tail regions, the distribution is equivalent to the Student's t-distribution with a direct mapping between q and the degrees of freedom. A practitioner using one of these distributions can therefore parameterize the same distribution in two different ways. The choice of the q-Gaussian form may arise if the system is non-extensive, or if there is lack of a connection to small samples sizes.

Characterization

[edit]

Probability density function

[edit]

The standard q-Gaussian has the probability density function [3]

f(x)=βCqeq(−βx2)

where

eq(x)=[1+(1−q)x]+11−q

is the q-exponential and the normalization factor Cq is given by

Cq=2πΓ(11−q)(3−q)1−qΓ(3−q2(1−q)) for −∞<q<1
Cq=π for q=1
Cq=πΓ(3−q2(q−1))q−1Γ(1q−1) for 1<q<3.

Note that for q<1 the q-Gaussian distribution is the PDF of a bounded random variable.

Cumulative density function

[edit]

For 1<q<3 cumulative density function is [5]

F(x)=12+q−1Γ(1q−1)xβ2F1(12,1q−1;32;−(q−1)βx2)πΓ(3−q2(q−1)),

where 2F1(a,b;c;z) is the hypergeometric function. As the hypergeometric function is defined for |z| < 1 but x is unbounded, Pfaff transformation could be used.

For q<1, F(x)={0x<−1β(1−q),12+1−qΓ(5−3q2(1−q))xβ2F1(12,1q−1;32;−(q−1)βx2)πΓ(2−q1−q)−1β(1−q)<x<1β(1−q),1x>1β(1−q).

Entropy

[edit]

Just as the normal distribution is the maximum information entropy distribution for fixed values of the first moment E⁡(X) and second moment E⁡(X2) (with the fixed zeroth moment E⁡(X0)=1 corresponding to the normalization condition), the q-Gaussian distribution is the maximum Tsallis entropy distribution for fixed values of these three moments.

[edit]

Student's t-distribution

[edit]

While it can be justified by an interesting alternative form of entropy, statistically it is a scaled reparametrization of the Student's t-distribution introduced by W. Gosset in 1908 to describe small-sample statistics. In Gosset's original presentation the degrees of freedom parameter ν was constrained to be a positive integer related to the sample size, but it is readily observed that Gosset's density function is valid for all real values of ν.[6] The scaled reparametrization introduces the alternative parameters q and β which are related to ν.

Given a Student's t-distribution with ν degrees of freedom, the equivalent q-Gaussian has

q=ν+3ν+1 with β=13−q

with inverse

ν=3−qq−1, but only if β=13−q.

Whenever β≠13−q, the function is simply a scaled version of Student's t-distribution.

It is sometimes argued that the distribution is a generalization of Student's t-distribution to negative and or non-integer degrees of freedom. However, the theory of Student's t-distribution extends trivially to all real degrees of freedom, where the support of the distribution is now compact rather than infinite in the case of ν < 0.

Three-parameter version

[edit]

As with many distributions centered on zero, the q-Gaussian can be trivially extended to include a location parameter μ. The density then becomes defined by

βCqeq(−β(x−μ)2).

Generating random deviates

[edit]

The Box–Muller transform has been generalized to allow random sampling from q-Gaussians.[7][8] The standard Box–Muller technique generates pairs of independent normally distributed variables from equations of the following form.

Z1=−2ln⁡(U1)cos⁡(2πU2)
Z2=−2ln⁡(U1)sin⁡(2πU2)

The generalized Box–Muller technique can generates pairs of q-Gaussian deviates that are not independent. In practice, only a single deviate will be generated from a pair of uniformly distributed variables. The following formula will generate deviates from a q-Gaussian with specified parameter q and β=13−q

Z=−2 lnq′(U1) cos(2πU2)

where  lnq is the q-logarithm and q′=1+q3−q

These deviates can be transformed to generate deviates from an arbitrary q-Gaussian by

Z′=μ+Zβ(3−q)

Applications

[edit]

Physics

[edit]

It has been shown that the momentum distribution of cold atoms in dissipative optical lattices is a q-Gaussian.[9]

The q-Gaussian distribution is also obtained as the asymptotic probability density function of the position of the unidimensional motion of a mass subject to two forces: a deterministic force of the type F1(x)=−2x/(1−x2) (determining an infinite potential well) and a stochastic white noise force F2(t)=2(1−q)ξ(t), where ξ(t) is a white noise. Note that in the overdamped/small mass approximation the above-mentioned convergence fails for q<0, as recently shown.[10]

Finance

[edit]

Financial return distributions in the New York Stock Exchange, NASDAQ and elsewhere have been interpreted as q-Gaussians.[11][12]

See also

[edit]

Notes

[edit]
  1. ↑ Tsallis, C. Nonadditive entropy and nonextensive statistical mechanics-an overview after 20 years. Braz. J. Phys. 2009, 39, 337–356
  2. ↑ d'Onofrio A. (ed.) Bounded Noises in Physics, Biology, and Engineering. Birkhauser (2013)
  3. 1 2 Umarov, Sabir; Tsallis, Constantino; Steinberg, Stanly (2008). "On a q-Central Limit Theorem Consistent with Nonextensive Statistical Mechanics" (PDF). Milan J. Math. 76. Birkhauser Verlag: 307–328. doi:10.1007/s00032-008-0087-y. S2CID 55967725. Retrieved 2011-07-27.
  4. ↑ Hilhorst, H.J. (2010), "Note on a q-modified central limit theorem", Journal of Statistical Mechanics: Theory and Experiment, 2010 (10) 10023, arXiv:1008.4259, Bibcode:2010JSMTE..10..023H, doi:10.1088/1742-5468/2010/10/P10023, S2CID 119316670.
  5. ↑ "TsallisQGaussianDistribution—Wolfram Language Documentation". reference.wolframcloud.com. Retrieved 2025-02-15.
  6. ↑ de Souza, André M. C.; Tsallis, Constantino (15 February 1997). "Student's t- and r-distributions: Unified derivation from an entropic variational principle". Physica A: Statistical Mechanics and Its Applications. 236 (1): 52–57. Bibcode:1997PhyA..236...52D. doi:10.1016/S0378-4371(96)00395-0.
  7. ↑ Thistleton, William J.; Marsh, John A.; Nelson, Kenric; Tsallis, Constantino (December 2007). "Generalized Box–MÜller Method for Generating $q$-Gaussian Random Deviates". IEEE Transactions on Information Theory. 53 (12): 4805–4810. doi:10.1109/TIT.2007.909173.
  8. ↑ Nelson, Kenric P.; Thistleton, William J. (October 2021). "Comments on "Generalized Box-Müller Method for Generating q -Gaussian Random Deviates"". IEEE Transactions on Information Theory. 67 (10): 6785–6789. Bibcode:2021ITIT...67.6785N. doi:10.1109/TIT.2021.3071489.
  9. ↑ Douglas, P.; Bergamini, S.; Renzoni, F. (2006). "Tunable Tsallis Distributions in Dissipative Optical Lattices" (PDF). Physical Review Letters. 96 (11) 110601. Bibcode:2006PhRvL..96k0601D. doi:10.1103/PhysRevLett.96.110601. PMID 16605807.
  10. ↑ Domingo, Dario; d'Onofrio, Alberto; Flandoli, Franco (2017). "Boundedness vs unboundedness of a noise linked to Tsallis q-statistics: The role of the overdamped approximation". Journal of Mathematical Physics. 58 (3). AIP Publishing: 033301. arXiv:1709.08260. Bibcode:2017JMP....58c3301D. doi:10.1063/1.4977081. ISSN 0022-2488. S2CID 84178785.
  11. ↑ Borland, Lisa (2002-08-07). "Option Pricing Formulas Based on a Non-Gaussian Stock Price Model". Physical Review Letters. 89 (9) 098701. American Physical Society (APS). arXiv:cond-mat/0204331. Bibcode:2002PhRvL..89i8701B. doi:10.1103/physrevlett.89.098701. ISSN 0031-9007. PMID 12190447. S2CID 5740827.
  12. ↑ L. Borland, The pricing of stock options, in Nonextensive Entropy – Interdisciplinary Applications, eds. M. Gell-Mann and C. Tsallis (Oxford University Press, New York, 2004)

Further reading

[edit]
[edit]