Period lattices of elliptic curves and related functions

Let \(E\) be an elliptic curve defined over a number field \(K\) (including \(\QQ\)). We attach a period lattice (a discrete rank 2 subgroup of \(\CC\)) to each embedding of \(K\) into \(\CC\).

In the case of real embeddings, the lattice is stable under complex conjugation and is called a real lattice. These have two types: rectangular, (the real curve has two connected components and positive discriminant) or non-rectangular (one connected component, negative discriminant).

The periods are computed to arbitrary precision using the AGM (Gauss’s Arithmetic-Geometric Mean).

EXAMPLES:

sage: K.<a> = NumberField(x^3-2)
sage: E = EllipticCurve([0,1,0,a,a])

First we try a real embedding:

sage: emb = K.embeddings(RealField())[0]
sage: L = E.period_lattice(emb); L
Period lattice associated to Elliptic Curve defined by y^2 = x^3 + x^2 + a*x + a over Number Field in a with defining polynomial x^3 - 2 with respect to the embedding Ring morphism:
From: Number Field in a with defining polynomial x^3 - 2
To:   Algebraic Real Field
Defn: a |--> 1.259921049894873?

The first basis period is real:

sage: L.basis()
(3.81452977217855, 1.90726488608927 + 1.34047785962440*I)
sage: L.is_real()
True

For a basis \(\omega_1,\omega_2\) normalised so that \(\omega_1/\omega_2\) is in the fundamental region of the upper half-plane, use the function normalised_basis() instead:

sage: L.normalised_basis()
(1.90726488608927 - 1.34047785962440*I, -1.90726488608927 - 1.34047785962440*I)

Next a complex embedding:

sage: emb = K.embeddings(ComplexField())[0]
sage: L = E.period_lattice(emb); L
Period lattice associated to Elliptic Curve defined by y^2 = x^3 + x^2 + a*x + a over Number Field in a with defining polynomial x^3 - 2 with respect to the embedding Ring morphism:
From: Number Field in a with defining polynomial x^3 - 2
To:   Algebraic Field
Defn: a |--> -0.6299605249474365? - 1.091123635971722?*I

In this case, the basis \(\omega_1\), \(\omega_2\) is always normalised so that \(\tau = \omega_1/\omega_2\) is in the fundamental region in the upper half plane:

sage: w1,w2 = L.basis(); w1,w2
(-1.37588604166076 - 2.58560946624443*I, -2.10339907847356 + 0.428378776460622*I)
sage: L.is_real()
False
sage: tau = w1/w2; tau
0.387694505032876 + 1.30821088214407*I
sage: L.normalised_basis()
(-1.37588604166076 - 2.58560946624443*I, -2.10339907847356 + 0.428378776460622*I)

We test that bug trac ticket #8415 (caused by a PARI bug fixed in v2.3.5) is OK:

sage: E = EllipticCurve('37a')
sage: K.<a> = QuadraticField(-7)
sage: EK = E.change_ring(K)
sage: EK.period_lattice(K.complex_embeddings()[0])
Period lattice associated to Elliptic Curve defined by y^2 + y = x^3 + (-1)*x over Number Field in a with defining polynomial x^2 + 7 with a = 2.645751311064591?*I with respect to the embedding Ring morphism:
  From: Number Field in a with defining polynomial x^2 + 7 with a = 2.645751311064591?*I
  To:   Algebraic Field
  Defn: a |--> -2.645751311064591?*I

REFERENCES:

• [CT2013]

AUTHORS:

• ?: initial version.

• John Cremona:

  • Adapted to handle real embeddings of number fields, September 2008.

  • Added basis_matrix function, November 2008

  • Added support for complex embeddings, May 2009.

  • Added complex elliptic logs, March 2010; enhanced, October 2010.

class sage.schemes.elliptic_curves.period_lattice.PeriodLattice(base_ring, rank, degree, sparse=False, coordinate_ring=None, category=None)

Bases: FreeModule_generic_pid

The class for the period lattice of an algebraic variety.

class sage.schemes.elliptic_curves.period_lattice.PeriodLattice_ell(E, embedding=None)

Bases: PeriodLattice

The class for the period lattice of an elliptic curve.

Currently supported are elliptic curves defined over \(\QQ\), and elliptic curves defined over a number field with a real or complex embedding, where the lattice constructed depends on that embedding.

basis(prec=None, algorithm='sage')

Return a basis for this period lattice as a 2-tuple.

INPUT:

• prec (default: None) – precision in bits (default precision if None).

• algorithm (string, default ‘sage’) – choice of implementation (for real embeddings only) between ‘sage’ (native Sage implementation) or ‘pari’ (use the PARI library: only available for real embeddings).

OUTPUT:

(tuple of Complex) \((\omega_1,\omega_2)\) where the lattice is \(\ZZ\omega_1 + \ZZ\omega_2\). If the lattice is real then \(\omega_1\) is real and positive, \(\Im(\omega_2)>0\) and \(\Re(\omega_1/\omega_2)\) is either \(0\) (for rectangular lattices) or \(\frac{1}{2}\) (for non-rectangular lattices). Otherwise, \(\omega_1/\omega_2\) is in the fundamental region of the upper half-plane. If the latter normalisation is required for real lattices, use the function normalised_basis() instead.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().basis()
(2.99345864623196, 2.45138938198679*I)

This shows that the issue reported at trac ticket #3954 is fixed:

sage: E = EllipticCurve('37a')
sage: b1 = E.period_lattice().basis(prec=30)
sage: b2 = E.period_lattice().basis(prec=30)
sage: b1 == b2
True

This shows that the issue reported at trac ticket #4064 is fixed:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().basis(prec=30)[0].parent()
Real Field with 30 bits of precision
sage: E.period_lattice().basis(prec=100)[0].parent()
Real Field with 100 bits of precision

sage: K.<a> = NumberField(x^3-2)
sage: emb = K.embeddings(RealField())[0]
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(emb)
sage: L.basis(64)
(3.81452977217854509, 1.90726488608927255 + 1.34047785962440202*I)

sage: emb = K.embeddings(ComplexField())[0]
sage: L = E.period_lattice(emb)
sage: w1,w2 = L.basis(); w1,w2
(-1.37588604166076 - 2.58560946624443*I, -2.10339907847356 + 0.428378776460622*I)
sage: L.is_real()
False
sage: tau = w1/w2; tau
0.387694505032876 + 1.30821088214407*I

basis_matrix(prec=None, normalised=False)

Return the basis matrix of this period lattice.

INPUT:

• prec (int or None``(default)) -- real precision in bits (default real precision if ``None).

• normalised (bool, default False) – if True and the embedding is real, use the normalised basis (see normalised_basis()) instead of the default.

OUTPUT:

A 2x2 real matrix whose rows are the lattice basis vectors, after identifying \(\CC\) with \(\RR^2\).

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().basis_matrix()
[ 2.99345864623196 0.000000000000000]
[0.000000000000000  2.45138938198679]

sage: K.<a> = NumberField(x^3-2)
sage: emb = K.embeddings(RealField())[0]
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(emb)
sage: L.basis_matrix(64)
[ 3.81452977217854509 0.000000000000000000]
[ 1.90726488608927255  1.34047785962440202]

See trac ticket #4388:

sage: L = EllipticCurve('11a1').period_lattice()
sage: L.basis_matrix()
[ 1.26920930427955 0.000000000000000]
[0.634604652139777  1.45881661693850]
sage: L.basis_matrix(normalised=True)
[0.634604652139777 -1.45881661693850]
[-1.26920930427955 0.000000000000000]

sage: L = EllipticCurve('389a1').period_lattice()
sage: L.basis_matrix()
[ 2.49021256085505 0.000000000000000]
[0.000000000000000  1.97173770155165]
sage: L.basis_matrix(normalised=True)
[ 2.49021256085505 0.000000000000000]
[0.000000000000000 -1.97173770155165]

complex_area(prec=None)

Return the area of a fundamental domain for the period lattice of the elliptic curve.

INPUT:

• prec (int or None``(default)) -- real precision in bits (default real precision if ``None).

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().complex_area()
7.33813274078958

sage: K.<a> = NumberField(x^3-2)
sage: embs = K.embeddings(ComplexField())
sage: E = EllipticCurve([0,1,0,a,a])
sage: [E.period_lattice(emb).is_real() for emb in K.embeddings(CC)]
[False, False, True]
sage: [E.period_lattice(emb).complex_area() for emb in embs]
[6.02796894766694, 6.02796894766694, 5.11329270448345]

coordinates(z, rounding=None)

Return the coordinates of a complex number w.r.t. the lattice basis

INPUT:

• z (complex) – A complex number.

• rounding (default None) – whether and how to round the

  output (see below).

OUTPUT:

When rounding is None (the default), returns a tuple of reals \(x\), \(y\) such that \(z=xw_1+yw_2\) where \(w_1\), \(w_2\) are a basis for the lattice (normalised in the case of complex embeddings).

When rounding is ‘round’, returns a tuple of integers \(n_1\), \(n_2\) which are the closest integers to the \(x\), \(y\) defined above. If \(z\) is in the lattice these are the coordinates of \(z\) with respect to the lattice basis.

When rounding is ‘floor’, returns a tuple of integers \(n_1\), \(n_2\) which are the integer parts to the \(x\), \(y\) defined above. These are used in reduce()

EXAMPLES:

sage: E = EllipticCurve('389a')
sage: L = E.period_lattice()
sage: w1, w2 = L.basis(prec=100)
sage: P = E([-1,1])
sage: zP = P.elliptic_logarithm(precision=100); zP
0.47934825019021931612953301006 + 0.98586885077582410221120384908*I
sage: L.coordinates(zP)
(0.19249290511394227352563996419, 0.50000000000000000000000000000)
sage: sum([x*w for x,w in zip(L.coordinates(zP), L.basis(prec=100))])
0.47934825019021931612953301006 + 0.98586885077582410221120384908*I

sage: L.coordinates(12*w1+23*w2)
(12.000000000000000000000000000, 23.000000000000000000000000000)
sage: L.coordinates(12*w1+23*w2, rounding='floor')
(11, 22)
sage: L.coordinates(12*w1+23*w2, rounding='round')
(12, 23)

curve()

Return the elliptic curve associated with this period lattice.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: L = E.period_lattice()
sage: L.curve() is E
True

sage: K.<a> = NumberField(x^3-2)
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(K.embeddings(RealField())[0])
sage: L.curve() is E
True

sage: L = E.period_lattice(K.embeddings(ComplexField())[0])
sage: L.curve() is E
True

e_log_RC(xP, yP, prec=None, reduce=True)

Return the elliptic logarithm of a real or complex point.

• xP, yP (real or complex) – Coordinates of a point on the embedded elliptic curve associated with this period lattice.

• prec (default: None) – real precision in bits (default real precision if None).

• reduce (default: True) – if True, the result is reduced with respect to the period lattice basis.

OUTPUT:

(complex number) The elliptic logarithm of the point \((xP,yP)\) with respect to this period lattice. If \(E\) is the elliptic curve and \(\sigma:K\to\CC\) the embedding, the returned value \(z\) is such that \(z\pmod{L}\) maps to \((xP,yP)=\sigma(P)\) under the standard Weierstrass isomorphism from \(\CC/L\) to \(\sigma(E)\). If reduce is True, the output is reduced so that it is in the fundamental period parallelogram with respect to the normalised lattice basis.

ALGORITHM:

Uses the complex AGM. See [CT2013] for details.

EXAMPLES:

sage: E = EllipticCurve('389a')
sage: L = E.period_lattice()
sage: P = E([-1,1])
sage: xP, yP = [RR(c) for c in P.xy()]

The elliptic log from the real coordinates:

sage: L.e_log_RC(xP, yP)
0.479348250190219 + 0.985868850775824*I

The same elliptic log from the algebraic point:

sage: L(P)
0.479348250190219 + 0.985868850775824*I

A number field example:

sage: K.<a> = NumberField(x^3-2)
sage: E = EllipticCurve([0,0,0,0,a])
sage: v = K.real_places()[0]
sage: L = E.period_lattice(v)
sage: P = E.lift_x(1/3*a^2 + a + 5/3)
sage: L(P)
3.51086196882538
sage: xP, yP = [v(c) for c in P.xy()]
sage: L.e_log_RC(xP, yP)
3.51086196882538

Elliptic logs of real points which do not come from algebraic points:

sage: ER = EllipticCurve([v(ai) for ai in E.a_invariants()])
sage: P = ER.lift_x(12.34)
sage: xP, yP = P.xy()
sage: xP, yP
(12.3400000000000, 43.3628968710567)
sage: L.e_log_RC(xP, yP)
3.76298229503967
sage: xP, yP = ER.lift_x(0).xy()
sage: L.e_log_RC(xP, yP)
2.69842609082114

Elliptic logs of complex points:

sage: v = K.complex_embeddings()[0]
sage: L = E.period_lattice(v)
sage: P = E.lift_x(1/3*a^2 + a + 5/3)
sage: L(P)
1.68207104397706 - 1.87873661686704*I
sage: xP, yP = [v(c) for c in P.xy()]
sage: L.e_log_RC(xP, yP)
1.68207104397706 - 1.87873661686704*I
sage: EC = EllipticCurve([v(ai) for ai in E.a_invariants()])
sage: xP, yP = EC.lift_x(0).xy()
sage: L.e_log_RC(xP, yP)
1.03355715602040 - 0.867257428417356*I

ei()

Return the x-coordinates of the 2-division points of the elliptic curve associated with this period lattice, as elements of QQbar.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: L = E.period_lattice()
sage: L.ei()
[-1.107159871688768?, 0.2695944364054446?, 0.8375654352833230?]

In the following example, we should have one purely real 2-division point coordinate, and two conjugate purely imaginary coordinates.

sage: K.<a> = NumberField(x^3-2)
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(K.embeddings(RealField())[0])
sage: x1,x2,x3 = L.ei()
sage: abs(x1.real())+abs(x2.real())<1e-14
True
sage: x1.imag(),x2.imag(),x3
(-1.122462048309373?, 1.122462048309373?, -1.000000000000000?)

sage: L = E.period_lattice(K.embeddings(ComplexField())[0])
sage: L.ei()
[-1.000000000000000? + 0.?e-1...*I,
-0.9720806486198328? - 0.561231024154687?*I,
0.9720806486198328? + 0.561231024154687?*I]

elliptic_exponential(z, to_curve=True)

Return the elliptic exponential of a complex number.

INPUT:

• z (complex) – A complex number (viewed modulo this period lattice).

• to_curve (bool, default True): see below.

OUTPUT:

• If to_curve is False, a 2-tuple of real or complex numbers representing the point \((x,y) = (\wp(z),\wp'(z))\) where \(\wp\) denotes the Weierstrass \(\wp\)-function with respect to this lattice.

• If to_curve is True, the point \((X,Y) = (x-b_2/12,y-(a_1(x-b_2/12)-a_3)/2)\) as a point in \(E(\RR)\) or \(E(\CC)\), with \((x,y) = (\wp(z),\wp'(z))\) as above, where \(E\) is the elliptic curve over \(\RR\) or \(\CC\) whose period lattice this is.

• If the lattice is real and \(z\) is also real then the output is a pair of real numbers if to_curve is True, or a point in \(E(\RR)\) if to_curve is False.

Note

The precision is taken from that of the input z.

EXAMPLES:

sage: E = EllipticCurve([1,1,1,-8,6])
sage: P = E(1,-2)
sage: L = E.period_lattice()
sage: z = L(P); z
1.17044757240090
sage: L.elliptic_exponential(z)
(0.999999999999999 : -2.00000000000000 : 1.00000000000000)
sage: _.curve()
Elliptic Curve defined by y^2 + 1.00000000000000*x*y + 1.00000000000000*y = x^3 + 1.00000000000000*x^2 - 8.00000000000000*x + 6.00000000000000 over Real Field with 53 bits of precision
sage: L.elliptic_exponential(z,to_curve=False)
(1.41666666666667, -2.00000000000000)
sage: z = L(P,prec=201); z
1.17044757240089592298992188482371493504472561677451007994189
sage: L.elliptic_exponential(z)
(1.00000000000000000000000000000000000000000000000000000000000 : -2.00000000000000000000000000000000000000000000000000000000000 : 1.00000000000000000000000000000000000000000000000000000000000)

Examples over number fields:

sage: x = polygen(QQ)
sage: K.<a> = NumberField(x^3-2)
sage: embs = K.embeddings(CC)
sage: E = EllipticCurve('37a')
sage: EK = E.change_ring(K)
sage: Li = [EK.period_lattice(e) for e in embs]
sage: P = EK(-1,-1)
sage: Q = EK(a-1,1-a^2)
sage: zi = [L.elliptic_logarithm(P) for L in Li]
sage: [c.real() for c in Li[0].elliptic_exponential(zi[0])]
[-1.00000000000000, -1.00000000000000, 1.00000000000000]
sage: [c.real() for c in Li[0].elliptic_exponential(zi[1])]
[-1.00000000000000, -1.00000000000000, 1.00000000000000]
sage: [c.real() for c in Li[0].elliptic_exponential(zi[2])]
[-1.00000000000000, -1.00000000000000, 1.00000000000000]

sage: zi = [L.elliptic_logarithm(Q) for L in Li]
sage: Li[0].elliptic_exponential(zi[0])
(-1.62996052494744 - 1.09112363597172*I : 1.79370052598410 - 1.37472963699860*I : 1.00000000000000)
sage: [embs[0](c) for c in Q]
[-1.62996052494744 - 1.09112363597172*I, 1.79370052598410 - 1.37472963699860*I, 1.00000000000000]
sage: Li[1].elliptic_exponential(zi[1])
(-1.62996052494744 + 1.09112363597172*I : 1.79370052598410 + 1.37472963699860*I : 1.00000000000000)
sage: [embs[1](c) for c in Q]
[-1.62996052494744 + 1.09112363597172*I, 1.79370052598410 + 1.37472963699860*I, 1.00000000000000]
sage: [c.real() for c in Li[2].elliptic_exponential(zi[2])]
[0.259921049894873, -0.587401051968199, 1.00000000000000]
sage: [embs[2](c) for c in Q]
[0.259921049894873, -0.587401051968200, 1.00000000000000]

Test to show that trac ticket #8820 is fixed:

sage: E = EllipticCurve('37a')
sage: K.<a> = QuadraticField(-5)
sage: L = E.change_ring(K).period_lattice(K.places()[0])
sage: L.elliptic_exponential(CDF(.1,.1))
(0.0000142854026029... - 49.9960001066650*I : 249.520141250950 + 250.019855549131*I : 1.00000000000000)
sage: L.elliptic_exponential(CDF(.1,.1), to_curve=False)
(0.0000142854026029447 - 49.9960001066650*I, 500.040282501900 + 500.039711098263*I)

\(z=0\) is treated as a special case:

sage: E = EllipticCurve([1,1,1,-8,6])
sage: L = E.period_lattice()
sage: L.elliptic_exponential(0)
(0.000000000000000 : 1.00000000000000 : 0.000000000000000)
sage: L.elliptic_exponential(0, to_curve=False)
(+infinity, +infinity)

sage: E = EllipticCurve('37a')
sage: K.<a> = QuadraticField(-5)
sage: L = E.change_ring(K).period_lattice(K.places()[0])
sage: P = L.elliptic_exponential(0); P
(0.000000000000000 : 1.00000000000000 : 0.000000000000000)
sage: P.parent()
Abelian group of points on Elliptic Curve defined by y^2 + 1.00000000000000*y = x^3 + (-1.00000000000000)*x over Complex Field with 53 bits of precision

Very small \(z\) are handled properly (see trac ticket #8820):

sage: K.<a> = QuadraticField(-1)
sage: E = EllipticCurve([0,0,0,a,0])
sage: L = E.period_lattice(K.complex_embeddings()[0])
sage: L.elliptic_exponential(1e-100)
(0.000000000000000 : 1.00000000000000 : 0.000000000000000)

The elliptic exponential of \(z\) is returned as (0 : 1 : 0) if the coordinates of z with respect to the period lattice are approximately integral:

sage: (100/log(2.0,10))/0.8
415.241011860920
sage: L.elliptic_exponential((RealField(415)(1e-100))).is_zero()
True
sage: L.elliptic_exponential((RealField(420)(1e-100))).is_zero()
False

elliptic_logarithm(P, prec=None, reduce=True)

Return the elliptic logarithm of a point.

INPUT:

• P (point) – A point on the elliptic curve associated with this period lattice.

• prec (default: None) – real precision in bits (default real precision if None).

• reduce (default: True) – if True, the result is reduced with respect to the period lattice basis.

OUTPUT:

(complex number) The elliptic logarithm of the point \(P\) with respect to this period lattice. If \(E\) is the elliptic curve and \(\sigma:K\to\CC\) the embedding, the returned value \(z\) is such that \(z\pmod{L}\) maps to \(\sigma(P)\) under the standard Weierstrass isomorphism from \(\CC/L\) to \(\sigma(E)\). If reduce is True, the output is reduced so that it is in the fundamental period parallelogram with respect to the normalised lattice basis.

ALGORITHM:

Uses the complex AGM. See [CT2013] for details.

EXAMPLES:

sage: E = EllipticCurve('389a')
sage: L = E.period_lattice()
sage: E.discriminant() > 0
True
sage: L.real_flag
1
sage: P = E([-1,1])
sage: P.is_on_identity_component ()
False
sage: L.elliptic_logarithm(P, prec=96)
0.4793482501902193161295330101 + 0.9858688507758241022112038491*I
sage: Q=E([3,5])
sage: Q.is_on_identity_component()
True
sage: L.elliptic_logarithm(Q, prec=96)
1.931128271542559442488585220

Note that this is actually the inverse of the Weierstrass isomorphism:

sage: L.elliptic_exponential(_)  # abs tol 1e-26
(3.000000000000000000000000000 : 5.000000000000000000000000000 : 1.000000000000000000000000000)

An example with negative discriminant, and a torsion point:

sage: E = EllipticCurve('11a1')
sage: L = E.period_lattice()
sage: E.discriminant() < 0
True
sage: L.real_flag
-1
sage: P = E([16,-61])
sage: L.elliptic_logarithm(P)
0.253841860855911
sage: L.real_period() / L.elliptic_logarithm(P)
5.00000000000000

An example where precision is problematic:

sage: E = EllipticCurve([1, 0, 1, -85357462, 303528987048]) #18074g1
sage: P = E([4458713781401/835903744, -64466909836503771/24167649046528, 1])
sage: L = E.period_lattice()
sage: L.ei()
[5334.003952567705? - 1.964393150436?e-6*I, 5334.003952567705? + 1.964393150436?e-6*I, -10668.25790513541?]
sage: L.elliptic_logarithm(P,prec=100)
0.27656204014107061464076203097

Some complex examples, taken from the paper by Cremona and Thongjunthug:

sage: K.<i> = QuadraticField(-1)
sage: a4 = 9*i-10
sage: a6 = 21-i
sage: E = EllipticCurve([0,0,0,a4,a6])
sage: e1 = 3-2*i; e2 = 1+i; e3 = -4+i
sage: emb = K.embeddings(CC)[1]
sage: L = E.period_lattice(emb)
sage: P = E(2-i,4+2*i)

By default, the output is reduced with respect to the normalised lattice basis, so that its coordinates with respect to that basis lie in the interval [0,1):

sage: z = L.elliptic_logarithm(P,prec=100); z
0.70448375537782208460499649302 - 0.79246725643650979858266018068*I
sage: L.coordinates(z)
(0.46247636364807931766105406092, 0.79497588726808704200760395829)

Using reduce=False this step can be omitted. In this case the coordinates are usually in the interval [-0.5,0.5), but this is not guaranteed. This option is mainly for testing purposes:

sage: z = L.elliptic_logarithm(P,prec=100, reduce=False); z
0.57002153834710752778063503023 + 0.46476340520469798857457031393*I
sage: L.coordinates(z)
(0.46247636364807931766105406092, -0.20502411273191295799239604171)

The elliptic logs of the 2-torsion points are half-periods:

sage: L.elliptic_logarithm(E(e1,0),prec=100)
0.64607575874356525952487867052 + 0.22379609053909448304176885364*I
sage: L.elliptic_logarithm(E(e2,0),prec=100)
0.71330686725892253793705940192 - 0.40481924028150941053684639367*I
sage: L.elliptic_logarithm(E(e3,0),prec=100)
0.067231108515357278412180731396 - 0.62861533082060389357861524731*I

We check this by doubling and seeing that the resulting coordinates are integers:

sage: L.coordinates(2*L.elliptic_logarithm(E(e1,0),prec=100))
(1.0000000000000000000000000000, 0.00000000000000000000000000000)
sage: L.coordinates(2*L.elliptic_logarithm(E(e2,0),prec=100))
(1.0000000000000000000000000000, 1.0000000000000000000000000000)
sage: L.coordinates(2*L.elliptic_logarithm(E(e3,0),prec=100))
(0.00000000000000000000000000000, 1.0000000000000000000000000000)

sage: a4 = -78*i + 104
sage: a6 = -216*i - 312
sage: E = EllipticCurve([0,0,0,a4,a6])
sage: emb = K.embeddings(CC)[1]
sage: L = E.period_lattice(emb)
sage: P = E(3+2*i,14-7*i)
sage: L.elliptic_logarithm(P)
0.297147783912228 - 0.546125549639461*I
sage: L.coordinates(L.elliptic_logarithm(P))
(0.628653378040238, 0.371417754610223)
sage: e1 = 1+3*i; e2 = -4-12*i; e3=-e1-e2
sage: L.coordinates(L.elliptic_logarithm(E(e1,0)))
(0.500000000000000, 0.500000000000000)
sage: L.coordinates(L.elliptic_logarithm(E(e2,0)))
(1.00000000000000, 0.500000000000000)
sage: L.coordinates(L.elliptic_logarithm(E(e3,0)))
(0.500000000000000, 0.000000000000000)

gens(prec=None, algorithm='sage')

Return a basis for this period lattice as a 2-tuple.

This is an alias for basis(). See the docstring there for a more in-depth explanation and further examples.

INPUT:

• prec (default: None) – precision in bits (default precision if None).

• algorithm (string, default ‘sage’) – choice of implementation (for real embeddings only) between ‘sage’ (native Sage implementation) or ‘pari’ (use the PARI library: only available for real embeddings).

OUTPUT:

(tuple of Complex) \((\omega_1,\omega_2)\) where the lattice is \(\ZZ\omega_1 + \ZZ\omega_2\). If the lattice is real then \(\omega_1\) is real and positive, \(\Im(\omega_2)>0\) and \(\Re(\omega_1/\omega_2)\) is either \(0\) (for rectangular lattices) or \(\frac{1}{2}\) (for non-rectangular lattices). Otherwise, \(\omega_1/\omega_2\) is in the fundamental region of the upper half-plane. If the latter normalisation is required for real lattices, use the function normalised_basis() instead.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().gens()
(2.99345864623196, 2.45138938198679*I)

sage: E.period_lattice().gens(prec = 100)
(2.9934586462319596298320099794, 2.4513893819867900608542248319*I)

is_real()

Return True if this period lattice is real.

EXAMPLES:

sage: f = EllipticCurve('11a')
sage: f.period_lattice().is_real()
True

sage: K.<i> = QuadraticField(-1)
sage: E = EllipticCurve(K,[0,0,0,i,2*i])
sage: emb = K.embeddings(ComplexField())[0]
sage: L = E.period_lattice(emb)
sage: L.is_real()
False

sage: K.<a> = NumberField(x^3-2)
sage: E = EllipticCurve([0,1,0,a,a])
sage: [E.period_lattice(emb).is_real() for emb in K.embeddings(CC)]
[False, False, True]

ALGORITHM:

The lattice is real if it is associated to a real embedding; such lattices are stable under conjugation.

is_rectangular()

Return True if this period lattice is rectangular.

Note

Only defined for real lattices; a RuntimeError is raised for non-real lattices.

EXAMPLES:

sage: f = EllipticCurve('11a')
sage: f.period_lattice().basis()
(1.26920930427955, 0.634604652139777 + 1.45881661693850*I)
sage: f.period_lattice().is_rectangular()
False

sage: f = EllipticCurve('37b')
sage: f.period_lattice().basis()
(1.08852159290423, 1.76761067023379*I)
sage: f.period_lattice().is_rectangular()
True

ALGORITHM:

The period lattice is rectangular precisely if the discriminant of the Weierstrass equation is positive, or equivalently if the number of real components is 2.

normalised_basis(prec=None, algorithm='sage')

Return a normalised basis for this period lattice as a 2-tuple.

INPUT:

• prec (default: None) – precision in bits (default precision if None).

• algorithm (string, default ‘sage’) – choice of implementation (for real embeddings only) between ‘sage’ (native Sage implementation) or ‘pari’ (use the PARI library: only available for real embeddings).

OUTPUT:

(tuple of Complex) \((\omega_1,\omega_2)\) where the lattice has the form \(\ZZ\omega_1 + \ZZ\omega_2\). The basis is normalised so that \(\omega_1/\omega_2\) is in the fundamental region of the upper half-plane. For an alternative normalisation for real lattices (with the first period real), use the function basis() instead.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().normalised_basis()
(2.99345864623196, -2.45138938198679*I)

sage: K.<a> = NumberField(x^3-2)
sage: emb = K.embeddings(RealField())[0]
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(emb)
sage: L.normalised_basis(64)
(1.90726488608927255 - 1.34047785962440202*I, -1.90726488608927255 - 1.34047785962440202*I)

sage: emb = K.embeddings(ComplexField())[0]
sage: L = E.period_lattice(emb)
sage: w1,w2 = L.normalised_basis(); w1,w2
(-1.37588604166076 - 2.58560946624443*I, -2.10339907847356 + 0.428378776460622*I)
sage: L.is_real()
False
sage: tau = w1/w2; tau
0.387694505032876 + 1.30821088214407*I

omega(prec=None, bsd_normalise=False)

Return the real or complex volume of this period lattice.

INPUT:

• prec (int or None``(default)) -- real precision in bits (default real precision if ``None)

• bsd_normalise (bool, default False) – flag to use BSD normalisation in the complex case.

OUTPUT:

(real) For real lattices, this is the real period times the number of connected components. For non-real lattices it is the complex area, or double the area if bsd_normalise is True.

Note

If the curve is given by a global minimal Weierstrass equation, then with bsd_normalise = True, this gives the correct period in the BSD conjecture: the product of this quantity over all embeddings appears in the BSD formula. In general a correction factor is required to make allowance for the model.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().omega()
5.98691729246392

This is not a minimal model:

sage: E = EllipticCurve([0,-432*6^2])
sage: E.period_lattice().omega()
0.486109385710056

If you were to plug the above omega into the BSD conjecture, you would get an incorrect value, out by a factor of 2. The following works though:

sage: F = E.minimal_model()
sage: F.period_lattice().omega()
0.972218771420113

sage: K.<a> = NumberField(x^3-2)
sage: emb = K.embeddings(RealField())[0]
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(emb)
sage: L.omega(64)
3.81452977217854509

A complex example (taken from J.E.Cremona and E.Whitley, Periods of cusp forms and elliptic curves over imaginary quadratic fields, Mathematics of Computation 62 No. 205 (1994), 407-429). See trac ticket #29645 and trac ticket #29782:

sage: K.<i> = QuadraticField(-1)
sage: E = EllipticCurve([0,1-i,i,-i,0])
sage: L = E.period_lattice(K.embeddings(CC)[0])
sage: L.omega()
8.80694160502647
sage: L.omega(prec=200)
8.8069416050264741493250743632295462227858630765392114070032
sage: L.omega(bsd_normalise=True)
17.6138832100529

real_period(prec=None, algorithm='sage')

Return the real period of this period lattice.

INPUT:

• prec (int or None (default)) – real precision in bits (default real precision if None)

• algorithm (string, default ‘sage’) – choice of implementation (for real embeddings only) between ‘sage’ (native Sage implementation) or ‘pari’ (use the PARI library: only available for real embeddings).

Note

Only defined for real lattices; a RuntimeError is raised for non-real lattices.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: E.period_lattice().real_period()
2.99345864623196

sage: K.<a> = NumberField(x^3-2)
sage: emb = K.embeddings(RealField())[0]
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(emb)
sage: L.real_period(64)
3.81452977217854509

reduce(z)

Reduce a complex number modulo the lattice

INPUT:

• z (complex) – A complex number.

OUTPUT:

(complex) the reduction of \(z\) modulo the lattice, lying in the fundamental period parallelogram with respect to the lattice basis. For curves defined over the reals (i.e. real embeddings) the output will be real when possible.

EXAMPLES:

sage: E = EllipticCurve('389a')
sage: L = E.period_lattice()
sage: w1, w2 = L.basis(prec=100)
sage: P = E([-1,1])
sage: zP = P.elliptic_logarithm(precision=100); zP
0.47934825019021931612953301006 + 0.98586885077582410221120384908*I
sage: z = zP+10*w1-20*w2; z
25.381473858740770069343110929 - 38.448885180257139986236950114*I
sage: L.reduce(z)
0.47934825019021931612953301006 + 0.98586885077582410221120384908*I
sage: L.elliptic_logarithm(2*P)
0.958696500380439
sage: L.reduce(L.elliptic_logarithm(2*P))
0.958696500380439
sage: L.reduce(L.elliptic_logarithm(2*P)+10*w1-20*w2)
0.958696500380444

sigma(z, prec=None, flag=0)

Return the value of the Weierstrass sigma function for this elliptic curve period lattice.

INPUT:

• z – a complex number

• prec (default: None) – real precision in bits

  (default real precision if None).

• flag –

  0: (default) ???;

  1: computes an arbitrary determination of log(sigma(z))

  2, 3: same using the product expansion instead of theta series. ???

Note

The reason for the ???’s above, is that the PARI documentation for ellsigma is very vague. Also this is only implemented for curves defined over \(\QQ\).

Todo

This function does not use any of the PeriodLattice functions and so should be moved to ell_rational_field.

EXAMPLES:

sage: EllipticCurve('389a1').period_lattice().sigma(CC(2,1))
2.60912163570108 - 0.200865080824587*I

tau(prec=None, algorithm='sage')

Return the upper half-plane parameter in the fundamental region.

INPUT:

• prec (default: None) – precision in bits (default precision if None).

• algorithm (string, default ‘sage’) – choice of implementation (for real embeddings only) between ‘sage’ (native Sage implementation) or ‘pari’ (use the PARI library: only available for real embeddings).

OUTPUT:

(Complex) \(\tau = \omega_1/\omega_2\) where the lattice has the form \(\ZZ\omega_1 + \ZZ\omega_2\), normalised so that \(\tau = \omega_1/\omega_2\) is in the fundamental region of the upper half-plane.

EXAMPLES:

sage: E = EllipticCurve('37a')
sage: L = E.period_lattice()
sage: L.tau()
1.22112736076463*I

sage: K.<a> = NumberField(x^3-2)
sage: emb = K.embeddings(RealField())[0]
sage: E = EllipticCurve([0,1,0,a,a])
sage: L = E.period_lattice(emb)
sage: tau = L.tau(); tau
-0.338718341018919 + 0.940887817679340*I
sage: tau.abs()
1.00000000000000
sage: -0.5 <= tau.real() <= 0.5
True

sage: emb = K.embeddings(ComplexField())[0]
sage: L = E.period_lattice(emb)
sage: tau = L.tau(); tau
0.387694505032876 + 1.30821088214407*I
sage: tau.abs()
1.36444961115933
sage: -0.5 <= tau.real() <= 0.5
True

sage.schemes.elliptic_curves.period_lattice.extended_agm_iteration(a, b, c)

Internal function for the extended AGM used in elliptic logarithm computation.

INPUT:

• a, b, c (real or complex) – three real or complex numbers.

OUTPUT:

(3-tuple) \((a_0,b_0,c_0)\), the limit of the iteration \((a,b,c) \mapsto ((a+b)/2,\sqrt{ab},(c+\sqrt(c^2+b^2-a^2))/2)\).

EXAMPLES:

sage: from sage.schemes.elliptic_curves.period_lattice import extended_agm_iteration
sage: extended_agm_iteration(RR(1),RR(2),RR(3))
(1.45679103104691, 1.45679103104691, 3.21245294970054)
sage: extended_agm_iteration(CC(1,2),CC(2,3),CC(3,4))
(1.46242448156430 + 2.47791311676267*I,
1.46242448156430 + 2.47791311676267*I,
3.22202144343535 + 4.28383734262540*I)

sage.schemes.elliptic_curves.period_lattice.normalise_periods(w1, w2)

Normalise the period basis \((w_1,w_2)\) so that \(w_1/w_2\) is in the fundamental region.

INPUT:

• w1,w2 (complex) – two complex numbers with non-real ratio

OUTPUT:

(tuple) \(((\omega_1',\omega_2'),[a,b,c,d])\) where \(a,b,c,d\) are integers such that

• \(ad-bc=\pm1\);

• \((\omega_1',\omega_2') = (a\omega_1+b\omega_2,c\omega_1+d\omega_2)\);

• \(\tau=\omega_1'/\omega_2'\) is in the upper half plane;

• \(|\tau|\ge1\) and \(|\Re(\tau)|\le\frac{1}{2}\).

EXAMPLES:

sage: from sage.schemes.elliptic_curves.period_lattice import reduce_tau, normalise_periods
sage: w1 = CC(1.234, 3.456)
sage: w2 = CC(1.234, 3.456000001)
sage: w1/w2    # in lower half plane!
0.999999999743367 - 9.16334785827644e-11*I
sage: w1w2, abcd = normalise_periods(w1,w2)
sage: a,b,c,d = abcd
sage: w1w2 == (a*w1+b*w2, c*w1+d*w2)
True
sage: w1w2[0]/w1w2[1]
1.23400010389203e9*I
sage: a*d-b*c # note change of orientation
-1

sage.schemes.elliptic_curves.period_lattice.reduce_tau(tau)

Transform a point in the upper half plane to the fundamental region.

INPUT:

• tau (complex) – a complex number with positive imaginary part

OUTPUT:

(tuple) \((\tau',[a,b,c,d])\) where \(a,b,c,d\) are integers such that

• \(ad-bc=1\);

• \(\tau`=(a\tau+b)/(c\tau+d)\);

• \(|\tau'|\ge1\);

• \(|\Re(\tau')|\le\frac{1}{2}\).

EXAMPLES:

sage: from sage.schemes.elliptic_curves.period_lattice import reduce_tau
sage: reduce_tau(CC(1.23,3.45))
(0.230000000000000 + 3.45000000000000*I, [1, -1, 0, 1])
sage: reduce_tau(CC(1.23,0.0345))
(-0.463960069171512 + 1.35591888067914*I, [-5, 6, 4, -5])
sage: reduce_tau(CC(1.23,0.0000345))
(0.130000000001761 + 2.89855072463768*I, [13, -16, 100, -123])