X-Git-Url: http://gitweb.michael.orlitzky.com/?a=blobdiff_plain;f=mjo%2Feja%2Feuclidean_jordan_algebra.py;h=a0ba1c68bb30bf54383807c52961ad82e4a69f03;hb=cc75c4e093a920c44f1eaaef4a1baa525b5c5727;hp=15ff26ca909cd1fbec4a2f1fc95e69b69f5bd02f;hpb=b82bd087507b8f727c10ca0eb55f9a1d15ed3438;p=sage.d.git diff --git a/mjo/eja/euclidean_jordan_algebra.py b/mjo/eja/euclidean_jordan_algebra.py index 15ff26c..a0ba1c6 100644 --- a/mjo/eja/euclidean_jordan_algebra.py +++ b/mjo/eja/euclidean_jordan_algebra.py @@ -71,6 +71,7 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): """ self._rank = rank self._natural_basis = natural_basis + self._multiplication_table = mult_table fda = super(FiniteDimensionalEuclideanJordanAlgebra, self) fda.__init__(field, mult_table, @@ -86,6 +87,102 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): return fmt.format(self.degree(), self.base_ring()) + + @cached_method + def _charpoly_coeff(self, i): + """ + Return the coefficient polynomial "a_{i}" of this algebra's + general characteristic polynomial. + + Having this be a separate cached method lets us compute and + store the trace/determinant (a_{r-1} and a_{0} respectively) + separate from the entire characteristic polynomial. + """ + (A_of_x, x) = self._charpoly_matrix() + R = A_of_x.base_ring() + A_cols = A_of_x.columns() + A_cols[i] = (x**self.rank()).vector() + numerator = column_matrix(A_of_x.base_ring(), A_cols).det() + denominator = A_of_x.det() + + # We're relying on the theory here to ensure that each a_i is + # indeed back in R, and the added negative signs are to make + # the whole charpoly expression sum to zero. + return R(-numerator/denominator) + + + @cached_method + def _charpoly_matrix(self): + """ + Compute the matrix whose entries A_ij are polynomials in + X1,...,XN. This same matrix is used in more than one method and + it's not so fast to construct. + """ + r = self.rank() + n = self.dimension() + + # Construct a new algebra over a multivariate polynomial ring... + names = ['X' + str(i) for i in range(1,n+1)] + R = PolynomialRing(self.base_ring(), names) + J = FiniteDimensionalEuclideanJordanAlgebra(R, + self._multiplication_table, + rank=r) + + idmat = identity_matrix(J.base_ring(), n) + + x = J(vector(R, R.gens())) + l1 = [column_matrix((x**k).vector()) for k in range(r)] + l2 = [idmat.column(k-1).column() for k in range(r+1, n+1)] + A_of_x = block_matrix(R, 1, n, (l1 + l2)) + return (A_of_x, x) + + + @cached_method + def characteristic_polynomial(self): + """ + EXAMPLES: + + The characteristic polynomial in the spin algebra is given in + Alizadeh, Example 11.11:: + + sage: J = JordanSpinEJA(3) + sage: p = J.characteristic_polynomial(); p + X1^2 - X2^2 - X3^2 + (-2*t)*X1 + t^2 + sage: xvec = J.one().vector() + sage: p(*xvec) + t^2 - 2*t + 1 + + """ + r = self.rank() + n = self.dimension() + + # The list of coefficient polynomials a_1, a_2, ..., a_n. + a = [ self._charpoly_coeff(i) for i in range(n) ] + + # We go to a bit of trouble here to reorder the + # indeterminates, so that it's easier to evaluate the + # characteristic polynomial at x's coordinates and get back + # something in terms of t, which is what we want. + R = a[0].parent() + S = PolynomialRing(self.base_ring(),'t') + t = S.gen(0) + S = PolynomialRing(S, R.variable_names()) + t = S(t) + + # Note: all entries past the rth should be zero. The + # coefficient of the highest power (x^r) is 1, but it doesn't + # appear in the solution vector which contains coefficients + # for the other powers (to make them sum to x^r). + if (r < n): + a[r] = 1 # corresponds to x^r + else: + # When the rank is equal to the dimension, trying to + # assign a[r] goes out-of-bounds. + a.append(1) # corresponds to x^r + + return sum( a[k]*(t**k) for k in range(len(a)) ) + + def inner_product(self, x, y): """ The inner product associated with this Euclidean Jordan algebra. @@ -266,19 +363,82 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): return A( (self.operator_matrix()**(n-1))*self.vector() ) + def apply_univariate_polynomial(self, p): + """ + Apply the univariate polynomial ``p`` to this element. + + A priori, SageMath won't allow us to apply a univariate + polynomial to an element of an EJA, because we don't know + that EJAs are rings (they are usually not associative). Of + course, we know that EJAs are power-associative, so the + operation is ultimately kosher. This function sidesteps + the CAS to get the answer we want and expect. + + EXAMPLES:: + + sage: R = PolynomialRing(QQ, 't') + sage: t = R.gen(0) + sage: p = t^4 - t^3 + 5*t - 2 + sage: J = RealCartesianProductEJA(5) + sage: J.one().apply_univariate_polynomial(p) == 3*J.one() + True + + TESTS: + + We should always get back an element of the algebra:: + + sage: set_random_seed() + sage: p = PolynomialRing(QQ, 't').random_element() + sage: J = random_eja() + sage: x = J.random_element() + sage: x.apply_univariate_polynomial(p) in J + True + + """ + if len(p.variables()) > 1: + raise ValueError("not a univariate polynomial") + P = self.parent() + R = P.base_ring() + # Convert the coeficcients to the parent's base ring, + # because a priori they might live in an (unnecessarily) + # larger ring for which P.sum() would fail below. + cs = [ R(c) for c in p.coefficients(sparse=False) ] + return P.sum( cs[k]*(self**k) for k in range(len(cs)) ) + + def characteristic_polynomial(self): """ - Return my characteristic polynomial (if I'm a regular - element). + Return the characteristic polynomial of this element. + + EXAMPLES: + + The rank of `R^3` is three, and the minimal polynomial of + the identity element is `(t-1)` from which it follows that + the characteristic polynomial should be `(t-1)^3`:: + + sage: J = RealCartesianProductEJA(3) + sage: J.one().characteristic_polynomial() + t^3 - 3*t^2 + 3*t - 1 + + Likewise, the characteristic of the zero element in the + rank-three algebra `R^{n}` should be `t^{3}`:: + + sage: J = RealCartesianProductEJA(3) + sage: J.zero().characteristic_polynomial() + t^3 + + The characteristic polynomial of an element should evaluate + to zero on that element:: + + sage: set_random_seed() + sage: x = RealCartesianProductEJA(3).random_element() + sage: p = x.characteristic_polynomial() + sage: x.apply_univariate_polynomial(p) + 0 - Eventually this should be implemented in terms of the parent - algebra's characteristic polynomial that works for ALL - elements. """ - if self.is_regular(): - return self.minimal_polynomial() - else: - raise NotImplementedError('irregular element') + p = self.parent().characteristic_polynomial() + return p(*self.vector()) def inner_product(self, other): @@ -492,8 +652,36 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): We can't use the superclass method because it relies on the algebra being associative. + + ALGORITHM: + + The usual way to do this is to check if the determinant is + zero, but we need the characteristic polynomial for the + determinant. The minimal polynomial is a lot easier to get, + so we use Corollary 2 in Chapter V of Koecher to check + whether or not the paren't algebra's zero element is a root + of this element's minimal polynomial. + + TESTS: + + The identity element is always invertible:: + + sage: set_random_seed() + sage: J = random_eja() + sage: J.one().is_invertible() + True + + The zero element is never invertible:: + + sage: set_random_seed() + sage: J = random_eja() + sage: J.zero().is_invertible() + False + """ - return not self.det().is_zero() + zero = self.parent().zero() + p = self.minimal_polynomial() + return not (p(zero) == zero) def is_nilpotent(self): @@ -598,14 +786,30 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): def minimal_polynomial(self): """ - EXAMPLES:: + Return the minimal polynomial of this element, + as a function of the variable `t`. + + ALGORITHM: + + We restrict ourselves to the associative subalgebra + generated by this element, and then return the minimal + polynomial of this element's operator matrix (in that + subalgebra). This works by Baes Proposition 2.3.16. + + TESTS: + + The minimal polynomial of the identity and zero elements are + always the same:: sage: set_random_seed() - sage: x = random_eja().random_element() - sage: x.degree() == x.minimal_polynomial().degree() - True + sage: J = random_eja() + sage: J.one().minimal_polynomial() + t - 1 + sage: J.zero().minimal_polynomial() + t - :: + The degree of an element is (by one definition) the degree + of its minimal polynomial:: sage: set_random_seed() sage: x = random_eja().random_element() @@ -626,31 +830,31 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): sage: y0 = y.vector()[0] sage: y_bar = y.vector()[1:] sage: actual = y.minimal_polynomial() - sage: x = SR.symbol('x', domain='real') - sage: expected = x^2 - 2*y0*x + (y0^2 - norm(y_bar)^2) + sage: t = PolynomialRing(J.base_ring(),'t').gen(0) + sage: expected = t^2 - 2*y0*t + (y0^2 - norm(y_bar)^2) sage: bool(actual == expected) True - """ - # The element we're going to call "minimal_polynomial()" on. - # Either myself, interpreted as an element of a finite- - # dimensional algebra, or an element of an associative - # subalgebra. - elt = None + The minimal polynomial should always kill its element:: - if self.parent().is_associative(): - elt = FiniteDimensionalAlgebraElement(self.parent(), self) - else: - V = self.span_of_powers() - assoc_subalg = self.subalgebra_generated_by() - # Mis-design warning: the basis used for span_of_powers() - # and subalgebra_generated_by() must be the same, and in - # the same order! - elt = assoc_subalg(V.coordinates(self.vector())) + sage: set_random_seed() + sage: x = random_eja().random_element() + sage: p = x.minimal_polynomial() + sage: x.apply_univariate_polynomial(p) + 0 - # Recursive call, but should work since elt lives in an - # associative algebra. - return elt.minimal_polynomial() + """ + V = self.span_of_powers() + assoc_subalg = self.subalgebra_generated_by() + # Mis-design warning: the basis used for span_of_powers() + # and subalgebra_generated_by() must be the same, and in + # the same order! + elt = assoc_subalg(V.coordinates(self.vector())) + + # We get back a symbolic polynomial in 'x' but want a real + # polynomial in 't'. + p_of_x = elt.operator_matrix().minimal_polynomial() + return p_of_x.change_variable_name('t') def natural_representation(self): @@ -853,7 +1057,10 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): # The dimension of the subalgebra can't be greater than # the big algebra, so just put everything into a list # and let span() get rid of the excess. - V = self.vector().parent() + # + # We do the extra ambient_vector_space() in case we're messing + # with polynomials and the direct parent is a module. + V = self.vector().parent().ambient_vector_space() return V.span( (self**d).vector() for d in xrange(V.dimension()) ) @@ -983,17 +1190,34 @@ class FiniteDimensionalEuclideanJordanAlgebra(FiniteDimensionalAlgebra): EXAMPLES:: sage: J = JordanSpinEJA(3) - sage: e0,e1,e2 = J.gens() - sage: x = e0 + e1 + e2 + sage: x = sum(J.gens()) sage: x.trace() 2 + :: + + sage: J = RealCartesianProductEJA(5) + sage: J.one().trace() + 5 + + TESTS: + + The trace of an element is a real number:: + + sage: set_random_seed() + sage: J = random_eja() + sage: J.random_element().trace() in J.base_ring() + True + """ - cs = self.characteristic_polynomial().coefficients(sparse=False) - if len(cs) >= 2: - return -1*cs[-2] - else: - raise ValueError('charpoly had fewer than 2 coefficients') + P = self.parent() + r = P.rank() + p = P._charpoly_coeff(r-1) + # The _charpoly_coeff function already adds the factor of + # -1 to ensure that _charpoly_coeff(r-1) is really what + # appears in front of t^{r-1} in the charpoly. However, + # we want the negative of THAT for the trace. + return -p(*self.vector()) def trace_inner_product(self, other): @@ -1084,12 +1308,17 @@ def random_eja(): Euclidean Jordan algebra of degree... """ - n = ZZ.random_element(1,5) - constructor = choice([RealCartesianProductEJA, - JordanSpinEJA, - RealSymmetricEJA, - ComplexHermitianEJA, - QuaternionHermitianEJA]) + + # The max_n component lets us choose different upper bounds on the + # value "n" that gets passed to the constructor. This is needed + # because e.g. R^{10} is reasonable to test, while the Hermitian + # 10-by-10 quaternion matrices are not. + (constructor, max_n) = choice([(RealCartesianProductEJA, 6), + (JordanSpinEJA, 6), + (RealSymmetricEJA, 5), + (ComplexHermitianEJA, 4), + (QuaternionHermitianEJA, 3)]) + n = ZZ.random_element(1, max_n) return constructor(n, field=QQ)