[go: up one dir, main page]

18 Orthogonal PolynomialsClassical Orthogonal Polynomials

§18.17 Integrals

Contents
  1. §18.17(i) Indefinite Integrals
  2. §18.17(ii) Integral Representations for Products
  3. §18.17(iii) Nicholson-Type Integrals
  4. §18.17(iv) Fractional Integrals
  5. §18.17(v) Fourier Transforms
  6. §18.17(vi) Laplace Transforms
  7. §18.17(vii) Mellin Transforms
  8. §18.17(viii) Other Integrals
  9. §18.17(ix) Compendia

§18.17(i) Indefinite Integrals

Jacobi

18.17.1 2⁢n⁢∫0x(1−y)α⁢(1+y)β⁢Pn(α,β)⁡(y)⁢dy=Pn−1(α+1,β+1)⁡(0)−(1−x)α+1⁢(1+x)β+1⁢Pn−1(α+1,β+1)⁡(x).

Laguerre

18.17.2 ∫0xLm⁡(y)⁢Ln⁡(x−y)⁢dy=∫0xLm+n⁡(y)⁢dy=Lm+n⁡(x)−Lm+n+1⁡(x).

Hermite

18.17.3 ∫0xHn⁡(y)⁢dy=12⁢(n+1)⁢(Hn+1⁡(x)−Hn+1⁡(0)),
18.17.4 ∫0xe−y2⁢Hn⁡(y)⁢dy=Hn−1⁡(0)−e−x2⁢Hn−1⁡(x).

Just as the indefinite integrals (18.17.1), (18.17.3) and (18.17.4), many similar formulas can be obtained by applying (1.4.26) to the differentiation formulas (18.9.15), (18.9.16) and (18.9.19)–(18.9.28).

§18.17(ii) Integral Representations for Products

Ultraspherical

18.17.5 Cn(λ)⁡(cos⁡θ1)Cn(λ)⁡(1)⁢Cn(λ)⁡(cos⁡θ2)Cn(λ)⁡(1)=Γ⁡(λ+12)π12⁢Γ⁡(λ)⁢∫0πCn(λ)⁡(cos⁡θ1⁢cos⁡θ2+sin⁡θ1⁢sin⁡θ2⁢cos⁡ϕ)Cn(λ)⁡(1)⁢(sin⁡ϕ)2⁢λ−1⁢dϕ,
λ>0.

Legendre

18.17.6 Pn⁡(cos⁡θ1)⁢Pn⁡(cos⁡θ2)=1π⁢∫0πPn⁡(cos⁡θ1⁢cos⁡θ2+sin⁡θ1⁢sin⁡θ2⁢cos⁡ϕ)⁢dϕ.

For formulas for Jacobi and Laguerre polynomials analogous to (18.17.5) and (18.17.6), see Koornwinder (1974, 1977). For addition formulas corresponding to (18.17.5) and (18.17.6) see (18.18.8) and (18.18.9), respectively.

§18.17(iii) Nicholson-Type Integrals

Legendre

For the Ferrers function 𝖰n⁡(x) and Legendre function Qn⁡(x) see §§14.3(i) and 14.3(ii), with μ=0 and ν=n.

Hermite

For the parabolic cylinder function V⁡(a,z) see §12.2. For similar formulas for ultraspherical polynomials see Durand (1975), and for Jacobi and Laguerre polynomials see Durand (1978).

§18.17(iv) Fractional Integrals

Jacobi

18.17.9 (1−x)α+μ⁢Pn(α+μ,β−μ)⁡(x)Γ⁡(α+μ+n+1)=∫x1(1−y)α⁢Pn(α,β)⁡(y)Γ⁡(α+n+1)⁢(y−x)μ−1Γ⁡(μ)⁢dy,
μ>0, −1<x<1,
18.17.10 xβ+μ⁢(x+1)nΓ⁡(β+μ+n+1)⁢Pn(α,β+μ)⁡(x−1x+1) =∫0xyβ⁢(y+1)nΓ⁡(β+n+1)⁢Pn(α,β)⁡(y−1y+1)⁢(x−y)μ−1Γ⁡(μ)⁢dy,
μ>0, x>0,
18.17.11 Γ⁡(n+α+β−μ+1)xn+α+β−μ+1⁢Pn(α,β−μ)⁡(1−2⁢x−1) =∫x∞Γ⁡(n+α+β+1)yn+α+β+1⁢Pn(α,β)⁡(1−2⁢y−1)⁢(y−x)μ−1Γ⁡(μ)⁢dy,
α+β+1>μ>0, x>1,

and three formulas similar to (18.17.9)–(18.17.11) by symmetry; compare the second row in Table 18.6.1. Formula (18.17.9), after substitution of (18.5.7), is a special case of (15.6.8). Formulas (18.17.9), (18.17.10) and (18.17.11) are fractional generalizations of n-th derivative formulas which are, after substitution of (18.5.7), special cases of (15.5.4), (15.5.5) and (15.5.3), respectively.

Ultraspherical

18.17.12 Γ⁡(λ−μ)⁢Cn(λ−μ)⁡(x−12)xλ−μ+12⁢n =∫x∞Γ⁡(λ)⁢Cn(λ)⁡(y−12)yλ+12⁢n⁢(y−x)μ−1Γ⁡(μ)⁢dy,
λ>μ>0, x>0,
18.17.13 x12⁢n⁢(x−1)λ+μ−12Γ⁡(λ+μ+12)⁢Cn(λ+μ)⁡(x−12)Cn(λ+μ)⁡(1) =∫1xy12⁢n⁢(y−1)λ−12Γ⁡(λ+12)⁢Cn(λ)⁡(y−12)Cn(λ)⁡(1)⁢(x−y)μ−1Γ⁡(μ)⁢dy,
μ>0, x>1.

Formulas (18.17.12) and (18.17.13) are fractional generalizations of the differentiation formulas given in Erdélyi et al. (1953b, §10.9(15)).

Laguerre

18.17.14 xα+μ⁢Ln(α+μ)⁡(x)Γ⁡(α+μ+n+1) =∫0xyα⁢Ln(α)⁡(y)Γ⁡(α+n+1)⁢(x−y)μ−1Γ⁡(μ)⁢dy,
μ>0, x>0.
18.17.15 e−x⁢Ln(α)⁡(x) =∫x∞e−y⁢Ln(α+μ)⁡(y)⁢(y−x)μ−1Γ⁡(μ)⁢dy,
μ>0.

Formulas (18.17.14) and (18.17.15) are fractional generalizations of n-th derivative formulas which are, after substitution of (13.6.19), special cases of (13.3.18) and (13.3.20), respectively.

§18.17(v) Fourier Transforms

Throughout this subsection we assume y>0; often however, this restriction can be eased by analytic continuation. In particular, in case of exponential Fourier transforms, we may assume y∈ℝ.

Jacobi

18.17.16 ∫−11(1−x)α⁢(1+x)β⁢Pn(α,β)⁡(x)⁢ei⁢x⁢y⁢dx=(i⁢y)n⁢ei⁢yn!⁢2n+α+β+1⁢B⁡(n+α+1,n+β+1)⁢F11⁡(n+α+1;2⁢n+α+β+2;−2⁢i⁢y).

For the beta function B⁡(a,b) see §5.12, and for the confluent hypergeometric function F11 see (16.2.1) and Chapter 13.

Ultraspherical

18.17.16_5 ∫−11(1−x2)λ−12⁢Cn(λ)⁡(x)⁢ei⁢x⁢y⁢dx=2⁢π⁢in⁢Γ⁡(n+2⁢λ)⁢Jn+λ⁡(y)n!⁢Γ⁡(λ)⁢(2⁢y)λ,
18.17.17 ∫01(1−x2)λ−12⁢C2⁢n(λ)⁡(x)⁢cos⁡(x⁢y)⁢dx=(−1)n⁢π⁢Γ⁡(2⁢n+2⁢λ)⁢Jλ+2⁢n⁡(y)(2⁢n)!⁢Γ⁡(λ)⁢(2⁢y)λ,
18.17.18 ∫01(1−x2)λ−12⁢C2⁢n+1(λ)⁡(x)⁢sin⁡(x⁢y)⁢dx=(−1)n⁢π⁢Γ⁡(2⁢n+2⁢λ+1)⁢J2⁢n+λ+1⁡(y)(2⁢n+1)!⁢Γ⁡(λ)⁢(2⁢y)λ.

For the Bessel function Jν see §10.2(ii).

Legendre

18.17.20 ∫01Pn⁡(1−2⁢x2)⁢cos⁡(x⁢y)⁢dx=(−1)n⁢12⁢π⁢Jn+12⁡(12⁢y)⁢J−n−12⁡(12⁢y),

Hermite

18.17.21_1 12⁢π⁢c⁢∫−∞∞e−12⁢x2/c⁢Hn⁡(x)⁢ei⁢x⁢y⁢dx=(i⁢2⁢c−1)n⁢e−12⁢c⁢y2⁢Hn⁡(c⁢y2⁢c−1),
ℜ⁡(c)>0, c≠12,

In (18.17.21_1) the branch choice of 2⁢c−1 for 0<c<12 is unimportant because on the right-hand side only even powers of 2⁢c−1 occur after expansion of the Hermite polynomial by (18.5.13). Formulas (18.17.21_2) and (18.17.21_3) are respectively the limit case c→12 and the special case c=1 of (18.17.21_1).

18.17.22 12⁢π⁢∫−∞∞e−14⁢x2⁢𝐻𝑒n⁡(x)⁢e12⁢i⁢x⁢y⁢dx=in⁢e−14⁢y2⁢𝐻𝑒n⁡(y),
18.17.23 ∫0∞e−12⁢x2⁢𝐻𝑒2⁢n⁡(x)⁢cos⁡(x⁢y)⁢dx=(−1)n⁢12⁢π⁢y2⁢n⁢e−12⁢y2,
18.17.24 ∫0∞e−x2⁢𝐻𝑒2⁢n⁡(2⁢x)⁢cos⁡(x⁢y)⁢dx=(−1)n⁢12⁢π⁢e−14⁢y2⁢𝐻𝑒2⁢n⁡(y).
18.17.26 ∫0∞e−12⁢x2⁢𝐻𝑒n⁡(x)⁢𝐻𝑒n+2⁢m+1⁡(x)⁢sin⁡(x⁢y)⁢dx=(−1)m⁢12⁢π⁢n!⁢y2⁢m+1⁢e−12⁢y2⁢Ln(2⁢m+1)⁡(y2).
18.17.27 ∫0∞e−12⁢x2⁢𝐻𝑒2⁢n+1⁡(x)⁢sin⁡(x⁢y)⁢dx=(−1)n⁢12⁢π⁢y2⁢n+1⁢e−12⁢y2,
18.17.28 ∫0∞e−x2⁢𝐻𝑒2⁢n+1⁡(2⁢x)⁢sin⁡(x⁢y)⁢dx=(−1)n⁢12⁢π⁢e−14⁢y2⁢𝐻𝑒2⁢n+1⁡(y).

Laguerre

18.17.28_5 ∫0∞e−x⁢xα⁢Ln(α)⁡(x)⁢ei⁢x⁢y⁢dx=Γ⁡(α+n+1)⁢(−i⁢y)nn!⁢(1−i⁢y)α+n+1,
18.17.30 ∫0∞x2⁢n⁢e−12⁢x2⁢Ln(n−12)⁡(12⁢x2)⁢cos⁡(x⁢y)⁢dx=12⁢π⁢y2⁢n⁢e−12⁢y2⁢Ln(n−12)⁡(12⁢y2),
18.17.31 ∫0∞e−a⁢x⁢xν−2⁢n⁢L2⁢n−1(ν−2⁢n)⁡(a⁢x)⁢cos⁡(x⁢y)⁢dx=i⁢(−1)n⁢Γ⁡(ν)2⁢(2⁢n−1)!⁢y2⁢n−1⁢((a+i⁢y)−ν−(a−i⁢y)−ν),
ν>2⁢n−1, a>0,
18.17.32 ∫0∞e−a⁢x⁢xν−1−2⁢n⁢L2⁢n(ν−1−2⁢n)⁡(a⁢x)⁢cos⁡(x⁢y)⁢dx=(−1)n⁢Γ⁡(ν)2⁢(2⁢n)!⁢y2⁢n⁢((a+i⁢y)−ν+(a−i⁢y)−ν),
ν>2⁢n, a>0.

§18.17(vi) Laplace Transforms

Many of the Fourier transforms given in §18.17(v) have analytic continuations to Laplace transforms. Some of the resulting formulas are given below.

Jacobi

18.17.33 ∫−11e−(x+1)⁢z⁢Pn(α,β)⁡(x)⁢(1−x)α⁢(1+x)β⁢dx=(−1)n⁢2α+β+n+1⁢Γ⁡(α+n+1)⁢Γ⁡(β+n+1)Γ⁡(α+β+2⁢n+2)⁢n!⁢zn⁢F11⁡(β+n+1α+β+2⁢n+2;−2⁢z),
z∈ℂ.

For the confluent hypergeometric function F11 see (16.2.1) and Chapter 13.

Laguerre

18.17.34_5 ∫0∞e−x⁢z⁢Lm(α)⁡(x)⁢Ln(α)⁡(x)⁢e−x⁢xα⁢dx=Γ⁡(α+m+1)⁢Γ⁡(α+n+1)Γ⁡(α+1)⁢m!⁢n!⁢zm+n(z+1)α+m+n+1⁢F12⁡(−m,−nα+1;z−2),
ℜ⁡z>−1.

Hermite

§18.17(vii) Mellin Transforms

Jacobi

18.17.36 ∫−11(1−x)z−1⁢(1+x)β⁢Pn(α,β)⁡(x)⁢dx=2β+z⁢Γ⁡(z)⁢Γ⁡(1+β+n)⁢(1+α−z)nn!⁢Γ⁡(1+β+z+n),
ℜ⁡z>0.

Ultraspherical

18.17.37 ∫01(1−x2)λ−12⁢Cn(λ)⁡(x)⁢xz−1⁢dx=π⁢ 21−2⁢λ−z⁢Γ⁡(n+2⁢λ)⁢Γ⁡(z)n!⁢Γ⁡(λ)⁢Γ⁡(12+12⁢n+λ+12⁢z)⁢Γ⁡(12+12⁢z−12⁢n),
ℜ⁡z>0.

Legendre

18.17.38 ∫01P2⁢n⁡(x)⁢xz−1⁢dx=(−1)n⁢(12−12⁢z)n2⁢(12⁢z)n+1,
ℜ⁡z>0,
18.17.39 ∫01P2⁢n+1⁡(x)⁢xz−1⁢dx=(−1)n⁢(1−12⁢z)n2⁢(12+12⁢z)n+1,
ℜ⁡z>−1.

Laguerre

This generalizes (18.17.34). For the hypergeometric function F12 see §§15.1 and 15.2(i).

Hermite

18.17.41 ∫0∞e−a⁢x⁢𝐻𝑒n⁡(x)⁢xz−1⁢dx=Γ⁡(z+n)⁢a−n−2⁢F22⁡(−12⁢n,−12⁢n+12−12⁢z−12⁢n,−12⁢z−12⁢n+12;−12⁢a2),
ℜ⁡a>0. Also, ℜ⁡z>0, n even; ℜ⁡z>−1, n odd.

For the generalized hypergeometric function F22 see (16.2.1).

§18.17(viii) Other Integrals

Ultraspherical

18.17.41_5 ∫−11Cℓ(λ)⁡(x)⁢Cm(λ)⁡(x)⁢Cn(λ)⁡(x)⁢(1−x2)λ−12⁢dx=(λ)12⁢ℓ+12⁢m−12⁢n⁢(λ)12⁢m+12⁢n−12⁢ℓ⁢(λ)12⁢n+12⁢ℓ−12⁢m⁢(2⁢λ)12⁢ℓ+12⁢m+12⁢n⁢Γ⁡(λ+12)⁢π(12⁢ℓ+12⁢m−12⁢n)!⁢(12⁢m+12⁢n−12⁢ℓ)!⁢(12⁢n+12⁢ℓ−12⁢m)!⁢Γ⁡(λ+12⁢ℓ+12⁢m+12⁢n+1),

provided that ℓ+m+n is even and the sum of any two of ℓ,m,n is not less than the third; otherwise the integral is zero.

Chebyshev

Legendre

18.17.44 ∫−11Pn⁡(x)−Pn⁡(t)|x−t|⁢dt=2⁢(1+12+⋯+1n)⁢Pn⁡(x),
−1≤x≤1.

The case x=1 is a limit case of an integral for Jacobi polynomials; see Askey and Razban (1972).

Laguerre

18.17.47 ∫0xtα⁢Lm(α)⁡(t)Lm(α)⁡(0)⁢(x−t)β⁢Ln(β)⁡(x−t)Ln(β)⁡(0)⁢dt=Γ⁡(α+1)⁢Γ⁡(β+1)Γ⁡(α+β+2)⁢xα+β+1⁢Lm+n(α+β+1)⁡(x)Lm+n(α+β+1)⁡(0).

Hermite

18.17.48 ∫−∞∞Hm⁡(y)⁢e−y2⁢Hn⁡(x−y)⁢e−(x−y)2⁢dy=π12⁢2−12⁢(m+n+1)⁢Hm+n⁡(2−12⁢x)⁢e−12⁢x2.
18.17.49 ∫−∞∞Hℓ⁡(x)⁢Hm⁡(x)⁢Hn⁡(x)⁢e−x2⁢dx=212⁢(ℓ+m+n)⁢ℓ!⁢m!⁢n!⁢π(12⁢ℓ+12⁢m−12⁢n)!⁢(12⁢m+12⁢n−12⁢ℓ)!⁢(12⁢n+12⁢ℓ−12⁢m)!,

provided that ℓ+m+n is even and the sum of any two of ℓ,m,n is not less than the third; otherwise the integral is zero.

Formulas (18.17.45) and (18.17.49) are integrated forms of the linearization formulas (18.18.22) and (18.18.23), respectively.

§18.17(ix) Compendia

For further integrals, see Apelblat (1983, pp. 189–204), Erdélyi et al. (1954a, pp. 38–39, 94–95, 170–176, 259–261, 324), Erdélyi et al. (1954b, pp. 42–44, 271–294), Gradshteyn and Ryzhik (2015, §§7.3–7.4), Gröbner and Hofreiter (1950, pp. 23–30), Marichev (1983, pp. 216–247), Oberhettinger (1972, pp. 64–67), Oberhettinger (1974, pp. 83–92), Oberhettinger (1990, pp. 44–47 and 152–154), Oberhettinger and Badii (1973, pp. 103–112), Prudnikov et al. (1986b, pp. 420–617), Prudnikov et al. (1992a, pp. 419–476), and Prudnikov et al. (1992b, pp. 280–308).