logoChat01.ai

Can you make a list of first spin-2 spherical harm...

Created on: July 31, 2025

Answered using O3 by Chat01

Question

Can you make a list of first spin-2 spherical harmonics up to l=3? What formula did you use to construct these functions?

Python

python
import sympy as sp

theta, phi = sp.symbols('theta phi', real=True)
l=3
m=2
s=2
# sympy wigner d function small d maybe sp.functions.special.tensor_functions.wigner_d?
dir(sp.functions.special)


['__builtins__',
 '__cached__',
 '__doc__',
 '__file__',
 '__loader__',
 '__name__',
 '__package__',
 '__path__',
 '__spec__',
 'bessel',
 'beta_functions',
 'bsplines',
 'delta_functions',
 'elliptic_integrals',
 'error_functions',
 'gamma_functions',
 'hyper',
 'mathieu_functions',
 'polynomials',
 'singularity_functions',
 'spherical_harmonics',
 'tensor_functions',
 'zeta_functions']

Python

python
dir(sp.functions.special.tensor_functions)


['Eijk',
 'Function',
 'Integer',
 'KroneckerDelta',
 'LeviCivita',
 'S',
 'SYMPY_INTS',
 '__builtins__',
 '__cached__',
 '__doc__',
 '__file__',
 '__loader__',
 '__name__',
 '__package__',
 '__spec__',
 'default_sort_key',
 'eval_levicivita',
 'fuzzy_not',
 'has_dups',
 'prod']

Python

python
dir(sp.functions.special.spherical_harmonics)


['Abs',
 'ArgumentIndexError',
 'Dummy',
 'Function',
 'I',
 'S',
 'Ynm',
 'Ynm_c',
 'Znm',
 '__builtins__',
 '__cached__',
 '__doc__',
 '__file__',
 '__loader__',
 '__name__',
 '__package__',
 '__spec__',
 '_x',
 'assoc_legendre',
 'cos',
 'cot',
 'exp',
 'factorial',
 'pi',
 'sin',
 'sqrt',
 'sympify']

Python

python
import sympy as sp
theta, phi = sp.symbols('theta phi', real=True)
def Y(l,m):
    return sp.functions.special.spherical_harmonics.Ynm(l,m,theta,phi) # complex spherical harmonic

def eth(f, s_in):
    sin = sp.sin(theta)
    return - (sin**s_in) * ( sp.diff( (sin**(-s_in)) * f, theta) + sp.I/sp.sin(theta) * sp.diff( (sin**(-s_in))*f, phi) )

def sY(l,m,s):
    # for s>=0 integer
    if s<0: raise NotImplementedError
    f = Y(l,m)
    s_in = 0
    for k in range(s):
        f = eth(f, s_in)
        s_in += 1
    norm = sp.sqrt(sp.factorial(l - s)/sp.factorial(l + s))
    return sp.simplify(norm * f)

sY22 = sp.simplify(sY(2,2,2))
sY21 = sp.simplify(sY(2,1,2))
sY20 = sp.simplify(sY(2,0,2))
sY2m1 = sp.simplify(sY(2,-1,2))
sY2m2 = sp.simplify(sY(2,-2,2))
sY22


sqrt(6)*(cos(theta) - 1)**2*Ynm(2, 2, theta, phi)/(6*sin(theta)**2)

Python

python
import sympy as sp
theta, phi = sp.symbols('theta phi', real=True)
def wigner_d(l,m,mp,theta):
    # small-d Wigner
    k=sp.symbols('k', integer=True)
    # pre-factor sqrt factorials
    pref = sp.sqrt(sp.factorial(l+m)*sp.factorial(l-m)*sp.factorial(l+mp)*sp.factorial(l-mp))
    expr = 0
    # determine k limits such that factorial args nonnegative
    # constraints: l+m-k>=0 => k<=l+m
    # k>=0
    # m'-m + k >= 0 => k >= m - m'
    # l - m' - k >=0 => k <= l - m'
    kmin = max(0, m - mp)
    kmax = min(l+m, l - mp)
    for ki in range(kmin, kmax+1):
        term = ((-1)**(ki-m+mp) * pref / (sp.factorial(l+m-ki)*sp.factorial(ki)*sp.factorial(mp-m+ki)*sp.factorial(l-mp-ki))
               * sp.cos(theta/2)**(2*l + m - mp - 2*ki)
               * sp.sin(theta/2)**(mp - m + 2*ki))
        expr += term
    return sp.simplify(expr)

def sY_by_wigner(l,m,s=2):
    d = wigner_d(l,m,-s,theta)
    return sp.simplify(sp.sqrt((2*l+1)/(4*sp.pi)) * d * sp.exp(sp.I*m*phi))

for L in (2,3):
    res = {}
    for m in range(-L, L+1):
        res[m] = sp.simplify(sp.together(sY_by_wigner(L,m,2)))
    print(L, res[2])
    print(L, res[1])
    print(L, res[0])

Python

python
from sympy import sin, cos, exp, I, pi, sqrt

def simplify_half_angle(expr):
    # Replace sin(θ/2)^odd * cos(θ/2)^odd patterns etc
    expr = sp.expand(expr)
    # We'll rewrite powers explicitly
    expr = sp.together(expr)
    expr = sp.simplify(expr)
    # Replace sin(θ/2)*cos(θ/2) -> sin θ /2
    expr = expr.replace(lambda x: x.is_Pow and x.base==sin(theta/2) and x.exp==1, sin(theta/2))
    expr = expr.replace(lambda x: x.is_Pow and x.base==cos(theta/2) and x.exp==1, cos(theta/2))
    # Use identities: sin(θ/2)^2 = (1 - cosθ)/2, cos(θ/2)^2 = (1 + cosθ)/2
    expr = expr.rewrite(sp.cos)
    expr = sp.simplify(expr)
    # Replace sin(θ/2)*cos(θ/2) with sinθ/2
    expr = expr.subs(sp.sin(theta/2)*sp.cos(theta/2), sp.sin(theta)/2)
    # Now replace even powers
    expr = expr.subs(sp.sin(theta/2)**2, (1-sp.cos(theta))/2)
    expr = expr.subs(sp.cos(theta/2)**2, (1+sp.cos(theta))/2)
    # Higher powers: reduce using above repeatedly
    expr = sp.simplify(sp.expand(expr))
    return expr

L=3
for m in range(-L,L+1):
    expr = sY_by_wigner(L,m,2)
    print(m, sp.simplify(sp.expand_trig(expr.rewrite(sp.cos))))

Python

python
import random
import cmath
import math

def sY2_formula_l2(m,theta,phi):
    if m==2:
        return math.sqrt(5/(64*math.pi))* (1-math.cos(theta))**2 * cmath.exp(2j*phi)
    if m==1:
        return math.sqrt(5/(16*math.pi))* math.sin(theta)*(1-math.cos(theta)) * cmath.exp(1j*phi)
    if m==0:
        return math.sqrt(15/(32*math.pi))* (math.sin(theta)**2)
    if m==-1:
        return -math.sqrt(5/(16*math.pi))* math.sin(theta)*(1+math.cos(theta)) * cmath.exp(-1j*phi)
    if m==-2:
        return math.sqrt(5/(64*math.pi))* (1+math.cos(theta))**2 * cmath.exp(-2j*phi)
    raise

def sY_wigner(l,m,theta,phi):
    # from earlier sympy result but numeric version using that function is heavy; let's just evaluate sympy expression we computed for given l,m
    expr = sY_by_wigner(l,m,2)
    f = sp.lambdify((theta,phi), expr, 'numpy')
    import numpy as np
    return complex(f(theta,phi))

theta_val = 1.1
phi_val = 0.7
for m in range(-2,3):
    a = sY2_formula_l2(m,theta_val,phi_val)
    b = sY_wigner(2,m,theta_val,phi_val)
    print(m, a, b, abs(a-b))

Python

python
def sY_wigner_numeric(l,m,theta_val,phi_val):
    expr = sY_by_wigner(l,m,2)
    f = sp.lambdify((theta,phi), expr, 'numpy')
    import numpy as np
    return complex(f(theta_val,phi_val))

theta_val = 1.1
phi_val = 0.7
for m in range(-2,3):
    a = sY2_formula_l2(m,theta_val,phi_val)
    b = sY_wigner_numeric(2,m,theta_val,phi_val)
    print(m, abs(a-b))

Python

python
def sY2_formula_l2_correct(m,theta,phi):
    if m==2:
        return math.sqrt(5/(64*math.pi))* (1-math.cos(theta))**2 * cmath.exp(2j*phi)
    if m==1:
        return math.sqrt(5/(16*math.pi))* math.sin(theta)*(1-math.cos(theta)) * cmath.exp(1j*phi)
    if m==0:
        return math.sqrt(15/(32*math.pi))* (math.sin(theta)**2)
    if m==-1:
        return math.sqrt(5/(16*math.pi))* math.sin(theta)*(1+math.cos(theta)) * cmath.exp(-1j*phi)
    if m==-2:
        return math.sqrt(5/(64*math.pi))* (1+math.cos(theta))**2 * cmath.exp(-2j*phi)
for m in range(-2,3):
    a = sY2_formula_l2_correct(m,theta_val,phi_val)
    b = sY_wigner_numeric(2,m,theta_val,phi_val)
    print(m, abs(a-b))

Python

python
def sY2_formula_l3(m,theta,phi):
    if m==3:
        return math.sqrt(42)/(16*math.sqrt(math.pi)) * math.sin(theta) * (1-math.cos(theta))**2 * cmath.exp(3j*phi)
    if m==2:
        return math.sqrt(7)/(8*math.sqrt(math.pi)) * (3*math.cos(theta)+2) * (1-math.cos(theta))**2 * cmath.exp(2j*phi)
    if m==1:
        return math.sqrt(70)/(16*math.sqrt(math.pi)) * (3*math.cos(theta)+1) * math.sin(theta) * (1-math.cos(theta)) * cmath.exp(1j*phi)
    if m==0:
        return math.sqrt(210)/(8*math.sqrt(math.pi))* (math.sin(theta)**2) * math.cos(theta)
    if m==-1:
        return math.sqrt(70)/(16*math.sqrt(math.pi)) * (3*math.cos(theta)-1) * math.sin(theta) * (1+math.cos(theta)) * cmath.exp(-1j*phi)
    if m==-2:
        return math.sqrt(7)/(8*math.sqrt(math.pi)) * (3*math.cos(theta)-2) * (1+math.cos(theta))**2 * cmath.exp(-2j*phi)
    if m==-3:
        return - math.sqrt(42)/(16*math.sqrt(math.pi)) * math.sin(theta) * (1+math.cos(theta))**2 * cmath.exp(-3j*phi)

for m in range(-3,4):
    a = sY2_formula_l3(m,theta_val,phi_val)
    b = sY_wigner_numeric(3,m,theta_val,phi_val)
    print(m, abs(a-b))

Thought

Formulas validated

Great! The formulas are now validated using the Wigner d approach. We can now present the normalized forms for l=2l = 2 and l=3l = 3, specifically for s=2s = 2.

For both cases, I’ll list the formulas as specified. I’ll also note the formula definition and state the normalization convention. In addition, I’ll clarify the use of the Condon-Shortley phase, domain conventions, and the symmetry relationships to avoid confusion. All constants and signs from the earlier derivation will be included.

Thought for 2m 59s

Answer

Below are the (spin-weight s = 2) spherical harmonics for the first allowed multipoles, ℓ = 2 and ℓ = 3. I use the standard Condon–Shortley phase and the usual normalization

∫sYℓm(θ,ϕ) sYℓ′m′∗(θ,ϕ) dΩ=δℓℓ′δmm′.\int {}_sY_{\ell m}(\theta,\phi)\,{}_sY^{*}_{\ell' m'}(\theta,\phi)\,d\Omega = \delta_{\ell\ell'}\delta_{mm'}.

Angles satisfy θ∈[0,π], ϕ∈[0,2π)\theta\in[0,\pi],\ \phi\in[0,2\pi).

Explicit functions

ℓ = 2

2Y2,2=564π (1−cos⁡θ)2 e2iϕ,2Y2,1=516π sin⁡θ (1−cos⁡θ) eiϕ,2Y2,0=1532π sin⁡2θ,2Y2,−1=516π sin⁡θ (1+cos⁡θ) e−iϕ,2Y2,−2=564π (1+cos⁡θ)2 e−2iϕ.\begin{aligned} {}_2Y_{2,2} &= \sqrt{\frac{5}{64\pi}}\,(1-\cos\theta)^2\,e^{2i\phi},\\[2pt] {}_2Y_{2,1} &= \sqrt{\frac{5}{16\pi}}\,\sin\theta\,(1-\cos\theta)\,e^{i\phi},\\[2pt] {}_2Y_{2,0} &= \sqrt{\frac{15}{32\pi}}\,\sin^2\theta,\\[2pt] {}_2Y_{2,-1}&= \sqrt{\frac{5}{16\pi}}\,\sin\theta\,(1+\cos\theta)\,e^{-i\phi},\\[2pt] {}_2Y_{2,-2}&= \sqrt{\frac{5}{64\pi}}\,(1+\cos\theta)^2\,e^{-2i\phi}. \end{aligned}

ℓ = 3

2Y3,3=4216π sin⁡θ (1−cos⁡θ)2 e3iϕ,2Y3,2=78π (3cos⁡θ+2) (1−cos⁡θ)2 e2iϕ,2Y3,1=7016π (3cos⁡θ+1) sin⁡θ (1−cos⁡θ) eiϕ,2Y3,0=2108π sin⁡2θ cos⁡θ,2Y3,−1=7016π (3cos⁡θ−1) sin⁡θ (1+cos⁡θ) e−iϕ,2Y3,−2=78π (3cos⁡θ−2) (1+cos⁡θ)2 e−2iϕ,2Y3,−3=− 4216π sin⁡θ (1+cos⁡θ)2 e−3iϕ.\begin{aligned} {}_2Y_{3,3} &= \frac{\sqrt{42}}{16\sqrt{\pi}}\,\sin\theta\,(1-\cos\theta)^2\,e^{3i\phi},\\[2pt] {}_2Y_{3,2} &= \frac{\sqrt{7}}{8\sqrt{\pi}}\,(3\cos\theta+2)\,(1-\cos\theta)^2\,e^{2i\phi},\\[2pt] {}_2Y_{3,1} &= \frac{\sqrt{70}}{16\sqrt{\pi}}\,(3\cos\theta+1)\,\sin\theta\,(1-\cos\theta)\,e^{i\phi},\\[2pt] {}_2Y_{3,0} &= \frac{\sqrt{210}}{8\sqrt{\pi}}\,\sin^2\theta\,\cos\theta,\\[2pt] {}_2Y_{3,-1} &= \frac{\sqrt{70}}{16\sqrt{\pi}}\,(3\cos\theta-1)\,\sin\theta\,(1+\cos\theta)\,e^{-i\phi},\\[2pt] {}_2Y_{3,-2} &= \frac{\sqrt{7}}{8\sqrt{\pi}}\,(3\cos\theta-2)\,(1+\cos\theta)^2\,e^{-2i\phi},\\[2pt] {}_2Y_{3,-3} &= -\,\frac{\sqrt{42}}{16\sqrt{\pi}}\,\sin\theta\,(1+\cos\theta)^2\,e^{-3i\phi}. \end{aligned}

(As a check, note the symmetry sYℓ,−m=(−1)s+m sYℓm∗ {}_sY_{\ell,-m}=(-1)^{s+m}\,{}_sY_{\ell m}^{*}.)

Formula used (two equivalent constructions)

(A) Wigner-d definition

sYℓm(θ,ϕ)=2ℓ+14π  dm,−s ℓ(θ)  eimϕ\boxed{\quad {}_sY_{\ell m}(\theta,\phi) = \sqrt{\frac{2\ell+1}{4\pi}}\; d^{\,\ell}_{m,-s}(\theta)\; e^{i m\phi}\quad}

where dmm′ ℓ(θ)d^{\,\ell}_{m m'}(\theta) is the (real) Wigner small-dd matrix element.
I generated the explicit forms above by evaluating dm,−2 ℓ(θ)d^{\,\ell}_{m,-2}(\theta) and simplifying.

(B) Spin-raising (eth) operator

ð sf=−sin⁡s ⁣θ(∂θ+i csc⁡θ ∂ϕ) ⁣[sin⁡−s ⁣θ sf],ð sYℓm=(ℓ−s)(ℓ+s+1)  s+1Yℓm.\eth\, {}_s f = -\sin^s\!\theta\left(\partial_\theta + i\,\csc\theta\,\partial_\phi\right)\!\left[\sin^{-s}\!\theta\, {}_s f\right],\qquad \eth\, {}_sY_{\ell m} = \sqrt{(\ell-s)(\ell+s+1)}\;{}_{s+1}Y_{\ell m}.

Starting from the scalar YℓmY_{\ell m} and applying ð\eth twice gives

2Yℓm=1ℓ(ℓ+1)(ℓ−1)(ℓ+2)  ð ð Yℓm\boxed{\quad {}_2Y_{\ell m} = \frac{1}{\sqrt{\ell(\ell+1)(\ell-1)(\ell+2)}}\;\eth\,\eth\,Y_{\ell m}\quad}

which yields the same expressions as above.

If you’d like, I can also list the s=−2s=-2 versions (often used in gravitational-wave work) or provide these in half-angle form.

Share this Q&A