Orthogonal Polynomials

• The Chebyshev polynomial of the first kind arises as a solution to the differential equation

  \[(1-x^2)\,y'' - x\,y' + n^2\,y = 0\]

  and those of the second kind as a solution to

  \[(1-x^2)\,y'' - 3x\,y' + n(n+2)\,y = 0.\]

  The Chebyshev polynomials of the first kind are defined by the recurrence relation

  \[T_0(x) = 1 \, T_1(x) = x \, T_{n+1}(x) = 2xT_n(x) - T_{n-1}(x). \,\]

  The Chebyshev polynomials of the second kind are defined by the recurrence relation

  \[U_0(x) = 1 \, U_1(x) = 2x \, U_{n+1}(x) = 2xU_n(x) - U_{n-1}(x). \,\]

  For integers \(m,n\), they satisfy the orthogonality relations

  \[\begin{split}\int_{-1}^1 T_n(x)T_m(x)\,\frac{dx}{\sqrt{1-x^2}} =\left\{ \begin{matrix} 0 &: n\ne m~~~~~\\ \pi &: n=m=0\\ \pi/2 &: n=m\ne 0 \end{matrix} \right.\end{split}\]

  and

  \[\int_{-1}^1 U_n(x)U_m(x)\sqrt{1-x^2}\,dx =\frac{\pi}{2}\delta_{m,n}.\]

  They are named after Pafnuty Chebyshev (alternative transliterations: Tchebyshef or Tschebyscheff).

• The Hermite polynomials are defined either by

  \[H_n(x)=(-1)^n e^{x^2/2}\frac{d^n}{dx^n}e^{-x^2/2}\]

  (the “probabilists’ Hermite polynomials”), or by

  \[H_n(x)=(-1)^n e^{x^2}\frac{d^n}{dx^n}e^{-x^2}\]

  (the “physicists’ Hermite polynomials”). Sage (via Maxima) implements the latter flavor. These satisfy the orthogonality relation

  \[\int_{-\infty}^\infty H_n(x)H_m(x)\,e^{-x^2}\,dx ={n!2^n}{\sqrt{\pi}}\delta_{nm}\]

  They are named in honor of Charles Hermite.

• Each Legendre polynomial \(P_n(x)\) is an \(n\)-th degree polynomial. It may be expressed using Rodrigues’ formula:

  \[P_n(x) = (2^n n!)^{-1} {\frac{d^n}{dx^n} } \left[ (x^2 -1)^n \right].\]

  These are solutions to Legendre’s differential equation:

  \[{\frac{d}{dx}} \left[ (1-x^2) {\frac{d}{dx}} P(x) \right] + n(n+1)P(x) = 0.\]

  and satisfy the orthogonality relation

  \[\int_{-1}^{1} P_m(x) P_n(x)\,dx = {\frac{2}{2n + 1}} \delta_{mn}\]

  The Legendre function of the second kind \(Q_n(x)\) is another (linearly independent) solution to the Legendre differential equation. It is not an “orthogonal polynomial” however.

  The associated Legendre functions of the first kind \(P_\ell^m(x)\) can be given in terms of the “usual” Legendre polynomials by

  \[\begin{split}\begin{array}{ll} P_\ell^m(x) &= (-1)^m(1-x^2)^{m/2}\frac{d^m}{dx^m}P_\ell(x) \\ &= \frac{(-1)^m}{2^\ell \ell!} (1-x^2)^{m/2}\frac{d^{\ell+m}}{dx^{\ell+m}}(x^2-1)^\ell. \end{array}\end{split}\]

  Assuming \(0 \le m \le \ell\), they satisfy the orthogonality relation:

  \[\int_{-1}^{1} P_k ^{(m)} P_\ell ^{(m)} dx = \frac{2 (\ell+m)!}{(2\ell+1)(\ell-m)!}\ \delta _{k,\ell},\]

  where \(\delta _{k,\ell}\) is the Kronecker delta.

  The associated Legendre functions of the second kind \(Q_\ell^m(x)\) can be given in terms of the “usual” Legendre polynomials by

  \[Q_\ell^m(x) = (-1)^m(1-x^2)^{m/2}\frac{d^m}{dx^m}Q_\ell(x).\]

  They are named after Adrien-Marie Legendre.

• Laguerre polynomials may be defined by the Rodrigues formula

  \[L_n(x)=\frac{e^x}{n!}\frac{d^n}{dx^n}\left(e^{-x} x^n\right).\]

  They are solutions of Laguerre’s equation:

  \[x\,y'' + (1 - x)\,y' + n\,y = 0\,\]

  and satisfy the orthogonality relation

  \[\int_0^\infty L_m(x) L_n(x) e^{-x}\,dx = \delta_{mn}.\]

  The generalized Laguerre polynomials may be defined by the Rodrigues formula:

  \[L_n^{(\alpha)}(x) = {\frac{x^{-\alpha} e^x}{n!}}{\frac{d^n}{dx^n}} \left(e^{-x} x^{n+\alpha}\right) .\]

  (These are also sometimes called the associated Laguerre polynomials.) The simple Laguerre polynomials are recovered from the generalized polynomials by setting \(\alpha =0\).

  They are named after Edmond Laguerre.

• Jacobi polynomials are a class of orthogonal polynomials. They are obtained from hypergeometric series in cases where the series is in fact finite:

  \[P_n^{(\alpha,\beta)}(z) =\frac{(\alpha+1)_n}{n!} \,_2F_1\left(-n,1+\alpha+\beta+n;\alpha+1;\frac{1-z}{2}\right) ,\]

  where \(()_n\) is Pochhammer’s symbol (for the rising factorial), (Abramowitz and Stegun p561.) and thus have the explicit expression

  \[P_n^{(\alpha,\beta)} (z) = \frac{\Gamma (\alpha+n+1)}{n!\Gamma (\alpha+\beta+n+1)} \sum_{m=0}^n \binom{n}{m} \frac{\Gamma (\alpha + \beta + n + m + 1)}{\Gamma (\alpha + m + 1)} \left(\frac{z-1}{2}\right)^m .\]

  They are named after Carl Jacobi.

• Ultraspherical or Gegenbauer polynomials are given in terms of the Jacobi polynomials \(P_n^{(\alpha,\beta)}(x)\) with \(\alpha=\beta=a-1/2\) by

  \[C_n^{(a)}(x)= \frac{\Gamma(a+1/2)}{\Gamma(2a)}\frac{\Gamma(n+2a)}{\Gamma(n+a+1/2)} P_n^{(a-1/2,a-1/2)}(x).\]

  They satisfy the orthogonality relation

  \[\int_{-1}^1(1-x^2)^{a-1/2}C_m^{(a)}(x)C_n^{(a)}(x)\, dx =\delta_{mn}2^{1-2a}\pi \frac{\Gamma(n+2a)}{(n+a)\Gamma^2(a)\Gamma(n+1)} ,\]

  for \(a>-1/2\). They are obtained from hypergeometric series in cases where the series is in fact finite:

  \[C_n^{(a)}(z) =\frac{(2a)^{\underline{n}}}{n!} \,_2F_1\left(-n,2a+n;a+\frac{1}{2};\frac{1-z}{2}\right)\]

  where \(\underline{n}\) is the falling factorial. (See Abramowitz and Stegun p561)

  They are named for Leopold Gegenbauer (1849-1903).

For completeness, the Pochhammer symbol, introduced by Leo August Pochhammer, \((x)_n\), is used in the theory of special functions to represent the “rising factorial” or “upper factorial”

\[(x)_n=x(x+1)(x+2)\cdots(x+n-1)=\frac{(x+n-1)!}{(x-1)!}.\]

On the other hand, the “falling factorial” or “lower factorial” is

\[x^{\underline{n}}=\frac{x!}{(x-n)!} ,\]

in the notation of Ronald L. Graham, Donald E. Knuth and Oren Patashnik in their book Concrete Mathematics.

Todo

Implement Zernike polynomials. Wikipedia article Zernike_polynomials

REFERENCES:

• [AS1964]

• Wikipedia article Chebyshev_polynomials

• Wikipedia article Legendre_polynomials

• Wikipedia article Hermite_polynomials

• http://mathworld.wolfram.com/GegenbauerPolynomial.html

• Wikipedia article Jacobi_polynomials

• Wikipedia article Laguerre_polynomia

• Wikipedia article Associated_Legendre_polynomials

• [Koe1999]

AUTHORS:

• David Joyner (2006-06)

• Stefan Reiterer (2010-)

• Ralf Stephan (2015-)

The original module wrapped some of the orthogonal/special functions in the Maxima package “orthopoly” and was written by Barton Willis of the University of Nebraska at Kearney.

class sage.functions.orthogonal_polys.ChebyshevFunction(name, nargs=2, latex_name=None, conversions={})

Bases: sage.functions.orthogonal_polys.OrthogonalFunction

Abstract base class for Chebyshev polynomials of the first and second kind.

EXAMPLES:

sage: chebyshev_T(3,x)
4*x^3 - 3*x

class sage.functions.orthogonal_polys.Func_assoc_legendre_P

Bases: sage.symbolic.function.BuiltinFunction

EXAMPLES:

sage: loads(dumps(gen_legendre_P))
gen_legendre_P
sage: maxima(gen_legendre_P(20,6,x, hold=True))._sage_().expand().coefficient(x,10)
2508866163428625/128

eval_poly(n, m, arg, **kwds)

Return the associated Legendre P(n, m, arg) polynomial for integers \(n > -1, m > -1\).

EXAMPLES:

sage: gen_legendre_P(7,4,x)
3465/2*(13*x^3 - 3*x)*(x^2 - 1)^2
sage: gen_legendre_P(3,1,sqrt(x))
-3/2*(5*x - 1)*sqrt(-x + 1)

REFERENCE:

• 20. 13. Dunster, Legendre and Related Functions, https://dlmf.nist.gov/14.7#E10

class sage.functions.orthogonal_polys.Func_assoc_legendre_Q

Bases: sage.symbolic.function.BuiltinFunction

EXAMPLES:

sage: loads(dumps(gen_legendre_Q))
gen_legendre_Q
sage: maxima(gen_legendre_Q(2,1,3, hold=True))._sage_().simplify_full()
1/4*sqrt(2)*(36*pi - 36*I*log(2) + 25*I)

eval_recursive(n, m, x, **kwds)

Return the associated Legendre Q(n, m, arg) function for integers \(n > -1, m > -1\).

EXAMPLES:

sage: gen_legendre_Q(3,4,x)
48/(x^2 - 1)^2
sage: gen_legendre_Q(4,5,x)
-384/((x^2 - 1)^2*sqrt(-x^2 + 1))
sage: gen_legendre_Q(0,1,x)
-1/sqrt(-x^2 + 1)
sage: gen_legendre_Q(0,2,x)
-1/2*((x + 1)^2 - (x - 1)^2)/(x^2 - 1)
sage: gen_legendre_Q(2,2,x).subs(x=2).expand()
9/2*I*pi - 9/2*log(3) + 14/3

class sage.functions.orthogonal_polys.Func_chebyshev_T

Bases: sage.functions.orthogonal_polys.ChebyshevFunction

Chebyshev polynomials of the first kind.

REFERENCE:

• [AS1964] 22.5.31 page 778 and 6.1.22 page 256.

EXAMPLES:

sage: chebyshev_T(5,x)
16*x^5 - 20*x^3 + 5*x
sage: var('k')
k
sage: test = chebyshev_T(k,x)
sage: test
chebyshev_T(k, x)

eval_algebraic(n, x)

Evaluate chebyshev_T as polynomial, using a recursive formula.

INPUT:

• n – an integer

• x – a value to evaluate the polynomial at (this can be any ring element)

EXAMPLES:

sage: chebyshev_T.eval_algebraic(5, x)
2*(2*(2*x^2 - 1)*x - x)*(2*x^2 - 1) - x
sage: chebyshev_T(-7, x) - chebyshev_T(7,x)
0
sage: R.<t> = ZZ[]
sage: chebyshev_T.eval_algebraic(-1, t)
t
sage: chebyshev_T.eval_algebraic(0, t)
1
sage: chebyshev_T.eval_algebraic(1, t)
t
sage: chebyshev_T(7^100, 1/2)
1/2
sage: chebyshev_T(7^100, Mod(2,3))
2
sage: n = 97; x = RIF(pi/2/n)
sage: chebyshev_T(n, cos(x)).contains_zero()
True
sage: R.<t> = Zp(2, 8, 'capped-abs')[]
sage: chebyshev_T(10^6+1, t)
(2^7 + O(2^8))*t^5 + O(2^8)*t^4 + (2^6 + O(2^8))*t^3 + O(2^8)*t^2 + (1 + 2^6 + O(2^8))*t + O(2^8)

eval_formula(n, x)

Evaluate chebyshev_T using an explicit formula. See [AS1964] 227 (p. 782) for details for the recursions. See also [Koe1999] for fast evaluation techniques.

INPUT:

• n – an integer

• x – a value to evaluate the polynomial at (this can be any ring element)

EXAMPLES:

sage: chebyshev_T.eval_formula(-1,x)
x
sage: chebyshev_T.eval_formula(0,x)
1
sage: chebyshev_T.eval_formula(1,x)
x
sage: chebyshev_T.eval_formula(2,0.1) == chebyshev_T._evalf_(2,0.1)
True
sage: chebyshev_T.eval_formula(10,x)
512*x^10 - 1280*x^8 + 1120*x^6 - 400*x^4 + 50*x^2 - 1
sage: chebyshev_T.eval_algebraic(10,x).expand()
512*x^10 - 1280*x^8 + 1120*x^6 - 400*x^4 + 50*x^2 - 1

class sage.functions.orthogonal_polys.Func_chebyshev_U

Bases: sage.functions.orthogonal_polys.ChebyshevFunction

Class for the Chebyshev polynomial of the second kind.

REFERENCE:

• [AS1964] 22.8.3 page 783 and 6.1.22 page 256.

EXAMPLES:

sage: R.<t> = QQ[]
sage: chebyshev_U(2,t)
4*t^2 - 1
sage: chebyshev_U(3,t)
8*t^3 - 4*t

eval_algebraic(n, x)

Evaluate chebyshev_U as polynomial, using a recursive formula.

INPUT:

• n – an integer

• x – a value to evaluate the polynomial at (this can be any ring element)

EXAMPLES:

sage: chebyshev_U.eval_algebraic(5,x)
-2*((2*x + 1)*(2*x - 1)*x - 4*(2*x^2 - 1)*x)*(2*x + 1)*(2*x - 1)
sage: parent(chebyshev_U(3, Mod(8,9)))
Ring of integers modulo 9
sage: parent(chebyshev_U(3, Mod(1,9)))
Ring of integers modulo 9
sage: chebyshev_U(-3,x) + chebyshev_U(1,x)
0
sage: chebyshev_U(-1,Mod(5,8))
0
sage: parent(chebyshev_U(-1,Mod(5,8)))
Ring of integers modulo 8
sage: R.<t> = ZZ[]
sage: chebyshev_U.eval_algebraic(-2, t)
-1
sage: chebyshev_U.eval_algebraic(-1, t)
0
sage: chebyshev_U.eval_algebraic(0, t)
1
sage: chebyshev_U.eval_algebraic(1, t)
2*t
sage: n = 97; x = RIF(pi/n)
sage: chebyshev_U(n-1, cos(x)).contains_zero()
True
sage: R.<t> = Zp(2, 6, 'capped-abs')[]
sage: chebyshev_U(10^6+1, t)
(2 + O(2^6))*t + O(2^6)

eval_formula(n, x)

Evaluate chebyshev_U using an explicit formula.

See [AS1964] 227 (p. 782) for details on the recursions. See also [Koe1999] for the recursion formulas.

INPUT:

• n – an integer

• x – a value to evaluate the polynomial at (this can be any ring element)

EXAMPLES:

sage: chebyshev_U.eval_formula(10, x)
1024*x^10 - 2304*x^8 + 1792*x^6 - 560*x^4 + 60*x^2 - 1
sage: chebyshev_U.eval_formula(-2, x)
-1
sage: chebyshev_U.eval_formula(-1, x)
0
sage: chebyshev_U.eval_formula(0, x)
1
sage: chebyshev_U.eval_formula(1, x)
2*x
sage: chebyshev_U.eval_formula(2,0.1) == chebyshev_U._evalf_(2,0.1)
True

class sage.functions.orthogonal_polys.Func_gen_laguerre

Bases: sage.functions.orthogonal_polys.OrthogonalFunction

REFERENCE:

• [AS1964] 22.5.16, page 778 and page 789.

class sage.functions.orthogonal_polys.Func_hermite

Bases: sage.symbolic.function.GinacFunction

Returns the Hermite polynomial for integers \(n > -1\).

REFERENCE:

• [AS1964] 22.5.40 and 22.5.41, page 779.

EXAMPLES:

sage: x = PolynomialRing(QQ, 'x').gen()
sage: hermite(2,x)
4*x^2 - 2
sage: hermite(3,x)
8*x^3 - 12*x
sage: hermite(3,2)
40
sage: S.<y> = PolynomialRing(RR)
sage: hermite(3,y)
8.00000000000000*y^3 - 12.0000000000000*y
sage: R.<x,y> = QQ[]
sage: hermite(3,y^2)
8*y^6 - 12*y^2
sage: w = var('w')
sage: hermite(3,2*w)
64*w^3 - 24*w
sage: hermite(5,3.1416)
5208.69733891963
sage: hermite(5,RealField(100)(pi))
5208.6167627118104649470287166

Check that trac ticket #17192 is fixed:

sage: x = PolynomialRing(QQ, 'x').gen()
sage: hermite(0,x)
1

sage: hermite(-1,x)
Traceback (most recent call last):
...
RuntimeError: hermite_eval: The index n must be a nonnegative integer

sage: hermite(-7,x)
Traceback (most recent call last):
...
RuntimeError: hermite_eval: The index n must be a nonnegative integer

sage: _ = var('m x')
sage: hermite(m, x).diff(m)
Traceback (most recent call last):
...
RuntimeError: derivative w.r.t. to the index is not supported yet

class sage.functions.orthogonal_polys.Func_jacobi_P

Bases: sage.functions.orthogonal_polys.OrthogonalFunction

Return the Jacobi polynomial \(P_n^{(a,b)}(x)\) for integers \(n > -1\) and a and b symbolic or \(a > -1\) and \(b > -1\). The Jacobi polynomials are actually defined for all a and b. However, the Jacobi polynomial weight \((1-x)^a(1+x)^b\) isn’t integrable for \(a \leq -1\) or \(b \leq -1\).

REFERENCE:

• Table on page 789 in [AS1964].

EXAMPLES:

sage: x = PolynomialRing(QQ, 'x').gen()
sage: jacobi_P(2,0,0,x)
3/2*x^2 - 1/2
sage: jacobi_P(2,1,2,1.2)
5.01000000000000

class sage.functions.orthogonal_polys.Func_laguerre

Bases: sage.functions.orthogonal_polys.OrthogonalFunction

REFERENCE:

• [AS1964] 22.5.16, page 778 and page 789.

class sage.functions.orthogonal_polys.Func_legendre_P

Bases: sage.symbolic.function.GinacFunction

EXAMPLES:

sage: legendre_P(4, 2.0)
55.3750000000000
sage: legendre_P(1, x)
x
sage: legendre_P(4, x+1)
35/8*(x + 1)^4 - 15/4*(x + 1)^2 + 3/8
sage: legendre_P(1/2, I+1.)
1.05338240025858 + 0.359890322109665*I
sage: legendre_P(0, SR(1)).parent()
Symbolic Ring

sage: legendre_P(0, 0)
1
sage: legendre_P(1, x)
x

sage: legendre_P(4, 2.)
55.3750000000000
sage: legendre_P(5.5,1.00001)
1.00017875754114
sage: legendre_P(1/2, I+1).n()
1.05338240025858 + 0.359890322109665*I
sage: legendre_P(1/2, I+1).n(59)
1.0533824002585801 + 0.35989032210966539*I
sage: legendre_P(42, RR(12345678))
2.66314881466753e309
sage: legendre_P(42, Reals(20)(12345678))
2.6632e309
sage: legendre_P(201/2, 0).n()
0.0561386178630179
sage: legendre_P(201/2, 0).n(100)
0.056138617863017877699963095883

sage: R.<x> = QQ[]
sage: legendre_P(4,x)
35/8*x^4 - 15/4*x^2 + 3/8
sage: legendre_P(10000,x).coefficient(x,1)
0
sage: var('t,x')
(t, x)
sage: legendre_P(-5,t)
35/8*t^4 - 15/4*t^2 + 3/8
sage: legendre_P(4, x+1)
35/8*(x + 1)^4 - 15/4*(x + 1)^2 + 3/8
sage: legendre_P(4, sqrt(2))
83/8
sage: legendre_P(4, I*e)
35/8*e^4 + 15/4*e^2 + 3/8

sage: n = var('n')
sage: derivative(legendre_P(n,x), x)
(n*x*legendre_P(n, x) - n*legendre_P(n - 1, x))/(x^2 - 1)
sage: derivative(legendre_P(3,x), x)
15/2*x^2 - 3/2
sage: derivative(legendre_P(n,x), n)
Traceback (most recent call last):
...
RuntimeError: derivative w.r.t. to the index is not supported yet

class sage.functions.orthogonal_polys.Func_legendre_Q

Bases: sage.symbolic.function.BuiltinFunction

EXAMPLES:

sage: loads(dumps(legendre_Q))
legendre_Q
sage: maxima(legendre_Q(20,x, hold=True))._sage_().coefficient(x,10)
-29113619535/131072*log(-(x + 1)/(x - 1))

eval_formula(n, arg, **kwds)

Return expanded Legendre Q(n, arg) function expression.

REFERENCE:

• 20. 13. Dunster, Legendre and Related Functions, https://dlmf.nist.gov/14.7#E2

EXAMPLES:

sage: legendre_Q.eval_formula(1, x)
1/2*x*(log(x + 1) - log(-x + 1)) - 1
sage: legendre_Q.eval_formula(2,x).expand().collect(log(1+x)).collect(log(1-x))
1/4*(3*x^2 - 1)*log(x + 1) - 1/4*(3*x^2 - 1)*log(-x + 1) - 3/2*x
sage: legendre_Q.eval_formula(20,x).coefficient(x,10)
-29113619535/131072*log(x + 1) + 29113619535/131072*log(-x + 1)
sage: legendre_Q(0, 2)
-1/2*I*pi + 1/2*log(3)
sage: legendre_Q(0, 2.)
0.549306144334055 - 1.57079632679490*I

eval_recursive(n, arg, **kwds)

Return expanded Legendre Q(n, arg) function expression.

EXAMPLES:

sage: legendre_Q.eval_recursive(2,x)
3/4*x^2*(log(x + 1) - log(-x + 1)) - 3/2*x - 1/4*log(x + 1) + 1/4*log(-x + 1)
sage: legendre_Q.eval_recursive(20,x).expand().coefficient(x,10)
-29113619535/131072*log(x + 1) + 29113619535/131072*log(-x + 1)

class sage.functions.orthogonal_polys.Func_ultraspherical

Bases: sage.symbolic.function.GinacFunction

Return the ultraspherical (or Gegenbauer) polynomial gegenbauer(n,a,x),

\[C_n^{a}(x)=\sum_{k=0}^{\lfloor n/2\rfloor} (-1)^k\frac{\Gamma(n-k+a)} {\Gamma(a)k!(n-2k)!}(2x)^{n-2k}.\]

When \(n\) is a nonnegative integer, this formula gives a polynomial in \(z\) of degree \(n\), but all parameters are permitted to be complex numbers. When \(a = 1/2\), the Gegenbauer polynomial reduces to a Legendre polynomial.

Computed using Pynac.

For numerical evaluation, consider using the mpmath library,, as it also allows complex numbers (and negative \(n\) as well); see the examples below.

REFERENCE:

• [AS1964] 22.5.27

EXAMPLES:

sage: ultraspherical(8, 101/11, x)
795972057547264/214358881*x^8 - 62604543852032/19487171*x^6...
sage: x = PolynomialRing(QQ, 'x').gen()
sage: ultraspherical(2,3/2,x)
15/2*x^2 - 3/2
sage: ultraspherical(1,1,x)
2*x
sage: t = PolynomialRing(RationalField(),"t").gen()
sage: gegenbauer(3,2,t)
32*t^3 - 12*t
sage: _ = var('x')
sage: for N in range(100):
....:     n = ZZ.random_element().abs() + 5
....:     a = QQ.random_element().abs() + 5
....:     assert ((n+1)*ultraspherical(n+1,a,x) - 2*x*(n+a)*ultraspherical(n,a,x) + (n+2*a-1)*ultraspherical(n-1,a,x)).expand().is_zero()
sage: ultraspherical(5,9/10,3.1416)
6949.55439044240
sage: ultraspherical(5,9/10,RealField(100)(pi))
6949.4695419382702451843080687

sage: _ = var('a n')
sage: gegenbauer(2,a,x)
2*(a + 1)*a*x^2 - a
sage: gegenbauer(3,a,x)
4/3*(a + 2)*(a + 1)*a*x^3 - 2*(a + 1)*a*x
sage: gegenbauer(3,a,x).expand()
4/3*a^3*x^3 + 4*a^2*x^3 + 8/3*a*x^3 - 2*a^2*x - 2*a*x
sage: gegenbauer(10,a,x).expand().coefficient(x,2)
1/12*a^6 + 5/4*a^5 + 85/12*a^4 + 75/4*a^3 + 137/6*a^2 + 10*a
sage: ex = gegenbauer(100,a,x)
sage: (ex.subs(a==55/98) - gegenbauer(100,55/98,x)).is_trivial_zero()
True

sage: gegenbauer(2,-3,x)
12*x^2 + 3
sage: gegenbauer(120,-99/2,3)
1654502372608570682112687530178328494861923493372493824
sage: gegenbauer(5,9/2,x)
21879/8*x^5 - 6435/4*x^3 + 1287/8*x
sage: gegenbauer(15,3/2,5)
3903412392243800

sage: derivative(gegenbauer(n,a,x),x)
2*a*gegenbauer(n - 1, a + 1, x)
sage: derivative(gegenbauer(3,a,x),x)
4*(a + 2)*(a + 1)*a*x^2 - 2*(a + 1)*a
sage: derivative(gegenbauer(n,a,x),a)
Traceback (most recent call last):
...
RuntimeError: derivative w.r.t. to the second index is not supported yet

Numerical evaluation with the mpmath library:

sage: from mpmath import gegenbauer as gegenbauer_mp
sage: from mpmath import mp
sage: mp.pretty = True; mp.dps=25
sage: gegenbauer_mp(-7,0.5,0.3)
0.1291811875
sage: gegenbauer_mp(2+3j, -0.75, -1000j)
(-5038991.358609026523401901 + 9414549.285447104177860806j)

class sage.functions.orthogonal_polys.OrthogonalFunction(name, nargs=2, latex_name=None, conversions={})

Bases: sage.symbolic.function.BuiltinFunction

Base class for orthogonal polynomials.

This class is an abstract base class for all orthogonal polynomials since they share similar properties. The evaluation as a polynomial is either done via maxima, or with pynac.

Convention: The first argument is always the order of the polynomial, the others are other values or parameters where the polynomial is evaluated.

eval_formula(*args)

Evaluate this polynomial using an explicit formula.

EXAMPLES:

sage: from sage.functions.orthogonal_polys import OrthogonalFunction
sage: P = OrthogonalFunction('testo_P')
sage: P.eval_formula(1,2.0)
Traceback (most recent call last):
...
NotImplementedError: no explicit calculation of values implemented