Useful Mathematical Formulae

Useful Mathematical Formulae

Appendix A Useful Mathematical Formulae x1−q − 1 ln x ≡ (x > 0, q ∈ R)(A.1) q 1 − q 1−q lnq x = x ln2−q x (x > 0; ∀q)(A.2) lnq (1/x) + ln2−q x = 0(x > 0; ∀q)(A.3) q q lnq x + ln(1/q)(1/x ) = 0(x > 0; ∀q)(A.4) ⎧ < < − / − , ⎪ 0ifq 1 and x 1 (1 q) ⎪ ⎪ ⎪ 1 ⎨⎪ + − 1−q < ≥− / − , 1 [1 (1 q) x] if q 1 and x 1 (1 q) x 1−q e ≡ [1 + (1 − q) x]+ ≡ q ⎪ ⎪ ex if q = 1(∀x) , ⎪ ⎪ ⎩ 1 [1 + (1 − q) x] 1−q if q > 1 and x < 1/(q − 1) . (A.5) x −x = ∀ eq e2−q 1(q)(A.6) x q −qx = ∀ (eq ) e(1/q) 1(q)(A.7) x+y+(1−q)xy = x y ∀ eq eq eq ( q)(A.8) 329 330 Appendix A Useful Mathematical Formulae x ⊕q y ≡ x + y + (1 − q) xy (A.9) For x ≥ 0 and y ≥ 0: ⎧ −q −q ⎪ 0ifq < 1andx 1 + y1 < 1 , ⎪ ⎪ ⎪ 1 ⎪ 1−q 1−q 1−q 1−q 1−q 1 ⎨ [x + y − 1] if q < 1andx + y ≥ 1 , 1−q 1−q 1−q x ⊗ y ≡ [x + y − 1]+ ≡ q ⎪ ⎪ xy if q = 1 ∀(x, y) , ⎪ ⎪ ⎩ − − 1 − − [x 1 q + y1 q − 1] 1−q if q > 1andx 1 q + y1 q > 1 . (A.10) 1 − x ⊗q y = [1 + (1 − q)(lnq x + lnq y)] 1 q (A.11) ⊕ x q y = x y ∀ eq eq eq ( q) (A.12) x+y = x ⊗ y ∀ eq eq q eq ( q) (A.13) d ln x 1 q = (x > 0; ∀q) (A.14) dx xq dex q = (ex )q (∀q) (A.15) dx q x q = qx ∀ (eq ) e2−(1/q) ( q) (A.16) x a = ax ∀ (eq ) e1−(1−q)/a ( q) (A.17) / − 2 − x 1/(q−1) 1 − b (q 1) a b b a− x x e = x q−1 e (b > 0; q > 1) (A.18) q q − 1 q Appendix A Useful Mathematical Formulae 331 1 1 3 1 2 1 ex = ex [1 − (1 − q)x2 + (1 − q)2x3(1 + x) − (1 − q)3x4(1 + x + x2) q 2 3 8 4 3 12 1 65 5 5 + (1 − q)4x5(1 + x + x2 + x3) 5 72 24 384 1 11 17 1 1 − (1 − q)5x6(1 + x + x2 + x3 + x4) + ...](q → 1; ∀x) 6 10 48 24 640 (A.19) 1 1 1 ln x = ln x [1 + (1 − q)lnx + (1 − q)2 ln2 x + (1 − q)3 ln3 x q 2 6 24 1 1 + (1 − q)4 ln4 x + (1 − q)5 ln5 x + ...](q → 1; x > 0) 120 720 (A.20) x ⊗q y = xy 1 − (1 − q)(ln x)(ln y) 1 + (1 − q)2 [(ln2 x)(ln y) + (ln x)(ln2 y) + (ln2 x)(ln2 y)] 2 1 − (1 − q)3 [2(ln3 x)(ln y) + 9(ln2 x)(ln2 y) + 2(ln x)(ln3 y) 12 +6(ln3 x)(ln2 y) + 6(ln2 x)(ln3 y) + 2(ln3 x)(ln3 y)] 1 + (1 − q)4 [(ln4 x)(ln y) + 14(ln3 x)(ln2 y) 24 +14(ln2 x)(ln3 y) + (ln x)(ln4 y) + 4 2 + 3 3 + 2 4 7(ln x)(ln y) 24(ln x)(ln y) 7(ln x)(ln y) +6(ln4 x)(ln3 y) + 6(ln3 x)(ln4 y) + (ln4 x)(ln4 y)] + ... (A.21) 1 1 1 ex = 1 + x + x2q + x3q(2q − 1) + x4q(2q − 1)(3q − 2) q 2 6 24 1 + x5q(2q − 1)(3q − 2)(4q − 3) + ... (x → 0; ∀q) (A.22) 120 1 1 1 ln (1 + x) = x − x2q + x3q(1 + q) − x4q(1 + q)(2 + q) q 2 6 24 1 + x5q(1 + q)(2 + q)(3 + q) + ... (x → 0; ∀q) (A.23) 120 x 1 b/(q−1)2 − q− 1 − a b b 1 a− x x e = x q−1 e (q > 1; b > 0) (A.24) q q − 1 q 332 Appendix A Useful Mathematical Formulae lnq x 2 lnq,q x ≡ lnq e (x > 0, (q, q ) ∈ R ) (A.25) 2 lnq,q (x ⊗q y) = lnq,q x ⊕q lnq,q y (x > 0, (q, q ) ∈ R ) (A.26) ∞ −β 1 1 2−q − α −α z = ααq−1 β(q−1) z eq 1 d e e (A.27) ⌫ 1 − q−1 [β(q − 1)] q 1 0 (α>0; β>0; 1 < q < 2) The following relations are useful for the Fourier transform of q-Gaussians (with β>0): ∞ eixp F (p) ≡ dx q + − β 2 1/(q−1) −∞ [1 (q 1) x ] ∞ cos (xp) = 2 dx 2 1/(q−1) 0 [1 + (q − 1) β x ] ⎧ √ 3−q ⎪ π 2−q 2 β(1−q) 2(1−q) p ⎪ ⌫ − √ < , ⎪ − β ( − ) J 3 q β − if q 1 ⎪ (1 q) 1 q p 1−q (1 q) ⎪ ⎪ ⎨ 2 π − p = e 4β if q = 1 , ⎪ β (A.28) ⎪ ⎪ ⎪ 3−q ⎪ π |p| 2(q−1) |p| ⎩ 2 √ − √ < < , β − β − K 3 q β − if 1 q 3 ⌫ 1 (q 1) 2 (q 1) 2(q−1) (q 1) q−1 where Jν (z) and Kν (z) are, respectively, the Bessel and the modified Bessel func- tions. For the three successive regions of q we have respectively used formulae 3.387-2 (page 346), 3.323-2 (page 333) and 8.432-5 (page 905) of [228] (see also [868]). For the q < 1 result we have taken into account the fact that the q-Gaussian |x| > √ 1 identically vanishes for β(1−q) . ∞ ∞ ξ ≡ iξ x ⊗ = iξ x[ f (x)]q−1 ≥ Fq [ f ]( ) dx eq q f (x) dx eq f (x)(q 1) (A.29) −∞ −∞ ∞ Fq [ f ](0) = dx f(x)(q ≥ 1) (A.30) −∞ ξ ∞ dFq [ f ]( ) q = i dx x [ f (x)] (q ≥ 1) (A.31) dξ ξ= −∞ 0 d2 F [ f ](ξ) ∞ q =−q dx x2 [ f (x)]2q−1 (q ≥ 1) (A.32) ξ 2 d ξ=0 −∞ Appendix A Useful Mathematical Formulae 333 d3 F [ f ](ξ) ∞ q =−iq(2q − 1) dx x3 [ f (x)]3q−2 (q ≥ 1) (A.33) ξ 3 d ξ=0 −∞ − d(n) F [ f ](ξ) n1 ∞ q = i n + m q − dx xn f x 1+n(q−1) n ( ) [1 ( 1)] [ ( )] dξ ξ= −∞ 0 m=0 (q ≥ 1; n = 1, 2, 3...) (A.34) 2−q Fq [af(ax)](ξ) = Fq [ f ](ξ/a )(a > 0; 1 ≤ q < 2) . (A.35) The generating function I (t)(t ∈ R) of a given distribution PN (N = 0, 1, 2, ...) is defined as follows: ∞ ∞ N I (t) ≡ t PN ( PN = 1) . (A.36) N=0 N=0 The negative binomial distribution is defined as follows: (N + k − 1)! N¯ /k N 1 k PN (N¯ , k) ≡ (N¯ > 0, k > 0) , (A.37) N!(k − 1)! 1 + N¯ /k 1 + N¯ /k where ∞ N¯ = NPN (N¯ , k) , (A.38) N=0 ∞ 1 [ (N − N¯ )2 P ] − N¯ = N=0 N . (A.39) k N¯ 2 Its generating function is given by = N¯ (t−1) , I (t) eq (A.40) with 1 q ≡ 1 + . (A.41) k The particular case q = 1 (i.e., k →∞) corresponds to the Poisson distribution 334 Appendix A Useful Mathematical Formulae N¯ N P (N¯ ) = e−N¯ , (A.42) N N! which satisfies the property that the width equals the mean value, i.e., ∞ 2 (N − N¯ ) PN = N¯ . (A.43) N=0 Appendix B Escort Distributions and q-Expectation Values B.1 First Example In order to illustrate the practical utility and peculiar properties of escort distribu- tions and their associated q-expectation values, we introduce and analyze here a pedagogical example [884].1 Let us assume that we have a set of empirical distributions { fn(x)} (n = 1, 2, 3, ...) defined as follows: A f (x) = n (λ>0; α ≥ 0) , (B.1) n (1 + λx)α if 0 ≤ x ≤ n, and zero otherwise. Normalization of fn(x) immediately yields λ(α − 1) A = . (B.2) n 1 − (1 + λ n)1−α In order to have finite values for An, ∀n, including n →∞(i.e., 0 < A∞ < ∞), α>1 is needed. Consequently A∞ = λ(α − 1) . (B.3) By identifying 1 α = , (B.4) q − 1 λ = β(q − 1) , (B.5) 1 The present illustration has greatly benefited from lengthy discussions with S. Abe, who launched [885] interesting questions regarding q-expectation values, and with E.M.F. Curado. 335 336 Appendix B Escort Distributions and q-Expectation Values Fig. B.1 The distributions fn (x)forn = 1, 2, 3, ∞ (from top to bottom) for (λ, α) = (2, 3/2) (from [884]). Equation (B.1) can be rewritten as = −β x β> ≥ . fn(x) An eq ( 0; q 1) (B.6) The variable x ≥ 0 could be a physical quantity, say earthquake intensity, mea- sured along small intervals, say 10−6, so small that sums can be replaced by integrals within an excellent approximation. The empiric distribution fn(x) could correspond to different seismic regions, say region 1 (for n = 1), region 2 (for n = 2), and so on. See Fig. B.1. Suppose we want to characterize the distribution fn(x) through its mean value. A straightforward calculation yields n 1 − (1 + λn)α + λn[α + (α − 1)λn] x (n) ≡ dx xf (x) = . (B.7) n α−1 0 (α − 2)λ(1 + λn)[1 − (1 + λn) ] This quantity is finite for all n (including n →∞)forα>2, but x (∞) diverges for 1 <α≤ 2.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    52 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us