def apply(self, z, evaluation):
'%(name)s[z__]'
args = z.get_sequence()
if len(args) != self.nargs:
return
# if no arguments are inexact attempt to use sympy
if all(not x.is_inexact() for x in args):
result = Expression(self.get_name(), *args).to_sympy()
result = self.prepare_mathics(result)
result = from_sympy(result)
# evaluate leaves to convert e.g. Plus[2, I] -> Complex[2, 1]
result = result.evaluate_leaves(evaluation)
else:
prec = min_prec(*args)
with mpmath.workprec(prec):
mpmath_args = [sympy2mpmath(x.to_sympy()) for x in args]
if None in mpmath_args:
return
try:
result = self.eval(*mpmath_args)
result = from_sympy(mpmath2sympy(result, prec))
except ValueError, exc:
text = str(exc)
if text == 'gamma function pole':
return Symbol('ComplexInfinity')
else:
raise
except ZeroDivisionError:
return
except SpecialValueError, exc:
return Symbol(exc.name)
def apply(self, z, evaluation):
"%(name)s[z__]"
args = z.get_sequence()
if len(args) != self.nargs:
return
# if no arguments are inexact attempt to use sympy
if len([True for x in args if Expression("InexactNumberQ", x).evaluate(evaluation).is_true()]) == 0:
expr = Expression(self.get_name(), *args).to_sympy()
result = from_sympy(expr)
# evaluate leaves to convert e.g. Plus[2, I] -> Complex[2, 1]
result = result.evaluate_leaves(evaluation)
else:
prec = min_prec(*args)
with mpmath.workprec(prec):
mpmath_args = [sympy2mpmath(x.to_sympy()) for x in args]
if None in mpmath_args:
return
try:
result = self.eval(*mpmath_args)
result = from_sympy(mpmath2sympy(result, prec))
except ValueError, exc:
text = str(exc)
if text == "gamma function pole":
return Symbol("ComplexInfinity")
else:
raise
except ZeroDivisionError:
return
except SpecialValueError, exc:
return Symbol(exc.name)
def apply(self, r, i, evaluation):
'Complex[r_?NumberQ, i_?NumberQ]'
if isinstance(r, Complex) or isinstance(i, Complex):
sym_form = r.to_sympy() + sympy.I * i.to_sympy()
r, i = sym_form.simplify().as_real_imag()
r, i = from_sympy(r), from_sympy(i)
return Complex(r, i)
def apply(self, items, evaluation):
'Times[items___]'
#TODO: Clean this up and optimise it
items = items.numerify(evaluation).get_sequence()
number = (sympy.Integer(1), sympy.Integer(0))
leaves = []
prec = min_prec(*items)
is_real = all([not isinstance(i, Complex) for i in items])
for item in items:
if isinstance(item, Number):
if isinstance(item, Complex):
sym_real, sym_imag = item.real.to_sympy(), item.imag.to_sympy()
else:
sym_real, sym_imag = item.to_sympy(), sympy.Integer(0)
if prec is not None:
sym_real = sym_real.n(dps(prec))
sym_imag = sym_imag.n(dps(prec))
if sym_real.is_zero and sym_imag.is_zero and prec is None:
return Integer('0')
number = (number[0]*sym_real - number[1]*sym_imag, number[0]*sym_imag + number[1]*sym_real)
elif leaves and item == leaves[-1]:
leaves[-1] = Expression('Power', leaves[-1], Integer(2))
elif leaves and item.has_form('Power', 2) and leaves[-1].has_form('Power', 2) and item.leaves[0].same(leaves[-1].leaves[0]):
leaves[-1].leaves[1] = Expression('Plus', item.leaves[1], leaves[-1].leaves[1])
elif leaves and item.has_form('Power', 2) and item.leaves[0].same(leaves[-1]):
leaves[-1] = Expression('Power', leaves[-1], Expression('Plus', item.leaves[1], Integer(1)))
elif leaves and leaves[-1].has_form('Power', 2) and leaves[-1].leaves[0].same(item):
leaves[-1] = Expression('Power', item, Expression('Plus', Integer(1), leaves[-1].leaves[1]))
else:
leaves.append(item)
if number == (1, 0):
number = None
elif number == (-1, 0) and leaves and leaves[0].has_form('Plus', None):
leaves[0].leaves = [Expression('Times', Integer(-1), leaf) for leaf in leaves[0].leaves]
number = None
if number is not None:
if number[1].is_zero and is_real:
leaves.insert(0, Number.from_mp(number[0], prec))
elif number[1].is_zero and number[1].is_Integer and prec is None:
leaves.insert(0, Number.from_mp(number[0], prec))
else:
leaves.insert(0, Complex(from_sympy(number[0]), from_sympy(number[1]), prec))
if not leaves:
return Integer(1)
elif len(leaves) == 1:
return leaves[0]
else:
return Expression('Times', *leaves)
def apply(self, m, evaluation):
'QRDecomposition[m_?MatrixQ]'
matrix = to_sympy_matrix(m)
try:
Q, R = matrix.QRdecomposition()
except sympy.matrices.MatrixError:
return evaluation.message('QRDecomposition', 'sympy')
Q = Q.transpose()
return Expression('List', *[from_sympy(Q), from_sympy(R)])
def append_last():
if last_item is not None:
if last_count == 1:
leaves.append(last_item)
else:
if last_item.has_form('Times', None):
last_item.leaves.insert(0, from_sympy(last_count))
leaves.append(last_item)
else:
leaves.append(Expression(
'Times', from_sympy(last_count), last_item))
def apply(self, m, l, evaluation):
'Norm[m_, l_]'
if isinstance(l, Symbol):
pass
elif isinstance(l, (Real, Integer)) and l.to_python() >= 1:
pass
else:
return evaluation.message('Norm', 'ptype', l)
l = l.to_sympy()
if l is None:
return
matrix = to_sympy_matrix(m)
if matrix is None:
return evaluation.message('Norm', 'nvm')
if len(matrix) == 0:
return
try:
res = matrix.norm(l)
except NotImplementedError:
return evaluation.message('Norm', 'normnotimplemented')
return from_sympy(res)
def apply(self, items, evaluation):
'Piecewise[items__]'
result = self.to_sympy(Expression('Piecewise', *items.get_sequence()))
if result is None:
return
if not isinstance(result, sympy.Piecewise):
return from_sympy(result)
def apply(self, expr, evaluation):
'Variables[expr_]'
variables = set()
def find_vars(e):
if e.to_sympy().is_constant():
return
elif e.is_symbol():
variables.add(e)
elif (e.has_form('Plus', None) or
e.has_form('Times', None)):
for l in e.leaves:
find_vars(l)
elif e.has_form('Power', 2):
(a, b) = e.leaves # a^b
if not(a.to_sympy().is_constant()) and b.to_sympy().is_rational:
find_vars(a)
elif not(e.is_atom()):
variables.add(e)
exprs = expr.leaves if expr.has_form('List', None) else [expr]
for e in exprs:
find_vars(from_sympy(e.to_sympy().expand()))
variables = Expression('List', *variables)
variables.sort() # MMA doesn't do this
return variables
def apply(self, expr, x, x0, evaluation, options={}):
'Limit[expr_, x_->x0_, OptionsPattern[Limit]]'
expr = expr.to_sympy()
x = x.to_sympy()
x0 = x0.to_sympy()
if expr is None or x is None or x0 is None:
return
direction = self.get_option(options, 'Direction', evaluation)
value = direction.get_int_value()
if value not in (-1, 1):
evaluation.message('Limit', 'ldir', direction)
if value > 0:
dir_sympy = '-'
else:
dir_sympy = '+'
try:
result = sympy.limit(expr, x, x0, dir_sympy)
except sympy.PoleError:
pass
except RuntimeError:
# Bug in Sympy: RuntimeError: maximum recursion depth exceeded
# while calling a Python object
pass
except NotImplementedError:
pass
except TypeError:
# Unknown SymPy0.7.6 bug
pass
else:
return from_sympy(result)
def apply(self, m, b, evaluation):
'LinearSolve[m_, b_]'
matrix = matrix_data(m)
if matrix is None:
return
if not b.has_form('List', None):
return
if len(b.leaves) != len(matrix):
return evaluation.message('LinearSolve', 'lslc')
system = [mm + [v] for mm, v in zip(matrix, b.leaves)]
system = to_sympy_matrix(system)
if system is None:
return
syms = [sympy.Dummy('LinearSolve_var%d' % k)
for k in range(system.cols - 1)]
sol = sympy.solve_linear_system(system, *syms)
if sol:
# substitute 0 for variables that are not in result dictionary
free_vars = dict((sym, sympy.Integer(
0)) for sym in syms if sym not in sol)
sol.update(free_vars)
sol = [(sol[sym] if sym in free_vars else sol[sym].subs(free_vars))
for sym in syms]
return from_sympy(sol)
else:
return evaluation.message('LinearSolve', 'nosol')
def apply(self, m, evaluation):
'Eigenvectors[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Eigenvectors', 'matsq', m)
# sympy raises an error for some matrices that Mathematica can compute.
try:
eigenvects = matrix.eigenvects()
except NotImplementedError:
return evaluation.message(
'Eigenvectors', 'eigenvecnotimplemented', m)
# The eigenvectors are given in the same order as the eigenvalues.
eigenvects = sorted(eigenvects, key=lambda (
val, c, vect): (abs(val), -val), reverse=True)
result = []
for val, count, basis in eigenvects:
# Select the i'th basis vector, convert matrix to vector,
# and convert from sympy
vects = [from_sympy(list(b)) for b in basis]
# This follows Mathematica convention better; higher indexed pivots
# are outputted first. e.g. {{0,1},{1,0}} instead of {{1,0},{0,1}}
vects.reverse()
# Add the vectors to results
result.extend(vects)
result.extend([Expression('List', *(
[0] * matrix.rows))] * (matrix.rows - len(result)))
return Expression('List', *result)
def apply(self, x, evaluation):
'Rationalize[x_]'
py_x = x.to_sympy()
if py_x is None or (not py_x.is_number) or (not py_x.is_real):
return x
return from_sympy(self.find_approximant(py_x))
def apply(self, expr, evaluation):
'Denominator[expr_]'
sympy_expr = expr.to_sympy()
if sympy_expr is None:
return None
numer, denom = sympy_expr.as_numer_denom()
return from_sympy(denom)
def apply(self, expr, evaluation):
'Simplify[expr_]'
sympy_expr = expr.to_sympy()
if sympy_expr is None:
return
sympy_result = sympy.simplify(sympy_expr)
return from_sympy(sympy_result)
def apply_exact(self, z, evaluation):
'%(name)s[z_?ExactNumberQ]'
expr = Expression(self.get_name(), z).to_sympy()
result = from_sympy(expr)
# evaluate leaves to convert e.g. Plus[2, I] -> Complex[2, 1]
result = result.evaluate_leaves(evaluation)
return result
def apply(self, m, evaluation):
'Det[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Det', 'matsq', m)
det = matrix.det()
return from_sympy(det)
def apply(self, m, evaluation):
'MatrixExp[m_]'
sympy_m = to_sympy_matrix(m)
try:
res = sympy_m.exp()
except NotImplementedError:
return evaluation.message('MatrixExp', 'matrixexpnotimplemented', m)
return from_sympy(res)
def apply(self, expr, evaluation):
'Together[expr_]'
expr_sympy = expr.to_sympy()
result = sympy.together(expr_sympy)
result = from_sympy(result)
result = cancel(result)
return result
def apply(self, m, power, evaluation):
'MatrixPower[m_, power_]'
sympy_m = to_sympy_matrix(m)
try:
res = sympy_m ** power.to_sympy()
except NotImplementedError:
return evaluation.message('MatrixPower', 'matrixpowernotimplemented', m)
return from_sympy(res)
def apply(self, m, evaluation):
'Inverse[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Inverse', 'matsq', m)
inv = matrix.inv()
return from_sympy(inv)
def apply(self, m, evaluation):
'RowReduce[m_]'
matrix = to_sympy_matrix(m)
if matrix is None:
return
reduced = matrix.rref()[0]
return from_sympy(reduced)
def apply(self, m, evaluation):
'Tr[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Tr', 'matsq', m)
tr = matrix.trace()
return from_sympy(tr)
def apply(self, m, evaluation):
'PseudoInverse[m_]'
matrix = to_sympy_matrix(m)
if matrix is None:
return evaluation.message('PseudoInverse', 'matrix', m, 1)
pinv = matrix.pinv()
return from_sympy(pinv)
def apply(self, a, b, evaluation):
'Cross[a_, b_]'
a = to_sympy_matrix(a)
b = to_sympy_matrix(b)
try:
res = a.cross(b)
except sympy.ShapeError:
return evaluation.message('Cross', 'nonn1')
return from_sympy(res)
def apply_N(self, k, precision, evaluation):
'N[AiryBiZero[k_Integer], precision_]'
prec = get_precision(precision, evaluation)
k_int = k.get_int_value()
with mpmath.workprec(prec):
result = mpmath2sympy(mpmath.airybizero(k_int), prec)
return from_sympy(result)
def apply(self, m, evaluation):
'NullSpace[m_]'
matrix = to_sympy_matrix(m)
if matrix is None:
return
nullspace = matrix.nullspace()
# convert n x 1 matrices to vectors
nullspace = [list(vec) for vec in nullspace]
return from_sympy(nullspace)
def apply(self, m, b, evaluation):
'LeastSquares[m_, b_]'
matrix = to_sympy_matrix(m)
b_vector = to_sympy_matrix([el.to_sympy() for el in b.leaves])
try:
solution = matrix.solve_least_squares(b_vector) # default method = Cholesky
except NotImplementedError as e:
return evaluation.message('LeastSquares', 'underdetermined')
return from_sympy(solution)
def apply(self, expr, var, evaluation):
'Apart[expr_, var_Symbol]'
expr_sympy = expr.to_sympy()
var_sympy = var.to_sympy()
try:
result = sympy.apart(expr_sympy, var_sympy)
result = from_sympy(result)
return result
except sympy.PolynomialError:
# raised e.g. for apart(sin(1/(x**2-y**2)))
return expr
def apply(self, m, evaluation):
'Eigenvalues[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Eigenvalues', 'matsq', m)
eigenvalues = matrix.eigenvals()
eigenvalues = sorted(eigenvalues.iteritems(), key=lambda (v, c): abs(v), reverse=True)
result = []
for val, count in eigenvalues:
result.extend([val] * count)
return from_sympy(result)
def apply(self, expr, var, evaluation):
"Apart[expr_, var_Symbol]"
expr_sympy = expr.to_sympy()
var_sympy = var.to_sympy()
if expr_sympy is None or var_sympy is None:
return None
try:
result = sympy.apart(expr_sympy, var_sympy)
result = from_sympy(result)
return result
except sympy.PolynomialError:
# raised e.g. for apart(sin(1/(x**2-y**2)))
return expr
def apply(self, expr, evaluation):
'Simplify[expr_]'
expr_sympy = expr.to_sympy()
result = expr_sympy
try:
result = sympy.simplify(result)
except TypeError:
# XXX What's going on here?
pass
result = sympy.trigsimp(result)
result = sympy.together(result)
result = sympy.cancel(result)
result = from_sympy(result)
return result
def apply(self, polys, xlist, alist, evaluation):
'ClassicalDixonResultant[polys_, xlist_,alist_]'
####this is not written in a wrapper function as it is not used anywhere else
res = DixonMatrix_apply(polys, xlist, alist)
if isinstance(res, String):
return res
if res.to_sympy() == sympy.sympify(0):
return res
rsym = res.leaves
dixon_list = [j.leaves for j in rsym]
dixon_sym = [[j.to_sympy() for j in i] for i in dixon_list]
dix_mat = sympy.Matrix(dixon_sym)
return from_sympy(sympy.det(dix_mat))
def apply(self, z, evaluation):
'%(name)s[z__]'
args = z.get_sequence()
if not self.check_nargs(len(args)):
return
# if no arguments are inexact attempt to use sympy
if all(not x.is_inexact() for x in args):
result = Expression(self.get_name(), *args).to_sympy()
result = self.prepare_mathics(result)
result = from_sympy(result)
# evaluate leaves to convert e.g. Plus[2, I] -> Complex[2, 1]
result = result.evaluate_leaves(evaluation)
else:
prec = min_prec(*args)
with mpmath.workprec(prec):
mpmath_args = [sympy2mpmath(x.to_sympy()) for x in args]
if None in mpmath_args:
return
try:
result = self.eval(*mpmath_args)
result = from_sympy(mpmath2sympy(result, prec))
except ValueError as exc:
text = str(exc)
if text == 'gamma function pole':
return Symbol('ComplexInfinity')
else:
raise
except ZeroDivisionError:
return
except SpecialValueError as exc:
return Symbol(exc.name)
return result
def apply(self, m, power, evaluation):
'MatrixPower[m_, power_]'
sympy_m = to_sympy_matrix(m)
if sympy_m is None:
return evaluation.message('MatrixPower', 'matrix', m, 1)
sympy_power = power.to_sympy()
if sympy_power is None:
return
try:
res = sympy_m ** sympy_power
except NotImplementedError:
return evaluation.message('MatrixPower', 'matrixpowernotimplemented', m)
return from_sympy(res)
def apply_constraints(self, f, vars, evaluation):
'Maximize[f_?ListQ, vars_]'
constraints = [function for function in f.leaves]
constraints[0] = from_sympy(constraints[0].to_sympy() * -1)
dual_solutions = Expression('Minimize', constraints,
vars).evaluate(evaluation).leaves
solutions = []
for dual_solution in dual_solutions:
solution_leaves = dual_solution.leaves
solutions.append([solution_leaves[0] * -1, solution_leaves[1]])
return from_python(solutions)
def apply(self, m, evaluation):
'Eigenvalues[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Eigenvalues', 'matsq', m)
eigenvalues = matrix.eigenvals()
try:
eigenvalues = sorted(eigenvalues.iteritems(),
key=lambda (v, c): (abs(v), -v), reverse=True)
except TypeError as e:
if not str(e).startswith('cannot determine truth value of'):
raise e
eigenvalues = eigenvalues.items()
return from_sympy([v for (v, c) in eigenvalues for _ in xrange(c)])
def apply(self, z, evaluation):
'%(name)s[z__]'
args = z.get_sequence()
if len(args) != self.nargs:
return
# if no arguments are inexact attempt to use sympy
if len([
True for x in args if Expression('InexactNumberQ', x).evaluate(
evaluation).is_true()
]) == 0:
expr = Expression(self.get_name(), *args).to_sympy()
result = from_sympy(expr)
# evaluate leaves to convert e.g. Plus[2, I] -> Complex[2, 1]
result = result.evaluate_leaves(evaluation)
else:
prec = min_prec(*args)
with mpmath.workprec(prec):
mpmath_args = [sympy2mpmath(x.to_sympy()) for x in args]
if None in mpmath_args:
return
try:
result = self.eval(*mpmath_args)
result = from_sympy(mpmath2sympy(result, prec))
except ValueError, exc:
text = str(exc)
if text == 'gamma function pole':
return Symbol('ComplexInfinity')
else:
raise
except ZeroDivisionError:
return
except SpecialValueError, exc:
return Symbol(exc.name)
def cancel(expr):
if expr.has_form('Plus', None):
return Expression('Plus', *[cancel(leaf) for leaf in expr.leaves])
else:
try:
result = expr.to_sympy()
# result = sympy.powsimp(result, deep=True)
result = sympy.cancel(result)
# cancel factors out rationals, so we factor them again
result = sympy_factor(result)
return from_sympy(result)
except sympy.PolynomialError:
# e.g. for non-commutative expressions
return expr
def apply_list(self, expr, vars, evaluation):
'FactorTermsList[expr_, vars_List]'
if expr == Integer(0):
return Expression('List', Integer(1), Integer(0))
elif isinstance(expr, Number):
return Expression('List', expr, Integer(1))
for x in vars.leaves:
if not(isinstance(x, Atom)):
return evaluation.message('CoefficientList', 'ivar', x)
sympy_expr = expr.to_sympy()
if sympy_expr is None:
return Expression('List', Integer(1), expr)
sympy_expr = sympy.together(sympy_expr)
sympy_vars = [x.to_sympy() for x in vars.leaves if isinstance(x, Symbol) and sympy_expr.is_polynomial(x.to_sympy())]
result = []
numer, denom = sympy_expr.as_numer_denom()
try:
from sympy import factor, factor_list, Poly
if denom == 1:
# Get numerical part
num_coeff, num_polys = factor_list(Poly(numer))
result.append(num_coeff)
# Get factors are independent of sub list of variables
if (sympy_vars and isinstance(expr, Expression)
and any(x.free_symbols.issubset(sympy_expr.free_symbols) for x in sympy_vars)):
for i in reversed(range(len(sympy_vars))):
numer = factor(numer) / factor(num_coeff)
num_coeff, num_polys = factor_list(Poly(numer), *[x for x in sympy_vars[:(i+1)]])
result.append(sympy.expand(num_coeff))
# Last factor
numer = factor(numer) / factor(num_coeff)
result.append(sympy.expand(numer))
else:
num_coeff, num_polys = factor_list(Poly(numer))
den_coeff, den_polys = factor_list(Poly(denom))
result = [num_coeff / den_coeff, sympy.expand(factor(numer)/num_coeff / (factor(denom)/den_coeff))]
except sympy.PolynomialError: # MMA does not raise error for non poly
result.append(sympy.expand(numer))
# evaluation.message(self.get_name(), 'poly', expr)
return Expression('List', *[from_sympy(i) for i in result])
def apply(self, f, xs, evaluation):
'Integrate[f_, xs__]'
f_sympy = f.to_sympy()
if isinstance(f_sympy, SympyExpression):
return
xs = xs.get_sequence()
vars = []
prec = None
for x in xs:
if x.has_form('List', 3):
x, a, b = x.leaves
prec_a = a.get_precision()
prec_b = b.get_precision()
if prec_a is not None and prec_b is not None:
prec_new = min(prec_a, prec_b)
if prec is None or prec_new < prec:
prec = prec_new
a = a.to_sympy()
b = b.to_sympy()
else:
a = b = None
a_mathics, b_mathics = a, b
if not x.get_name():
evaluation.message('Integrate', 'ilim')
return
x = x.to_sympy()
if a is None or b is None:
vars.append(x)
else:
vars.append((x, a, b))
try:
result = sympy.integrate(f_sympy, *vars)
except sympy.PolynomialError:
return
except ValueError:
# e.g. ValueError: can't raise polynomial to a negative power
return
except NotImplementedError:
# e.g. NotImplementedError: Result depends on the sign of
# -sign(_Mathics_User_j)*sign(_Mathics_User_w)
return
if prec is not None:
result = sympy.N(result)
result = from_sympy(result)
return result
def apply_iter(self, expr, i, imin, imax, di, evaluation):
'%(name)s[expr_, {i_Symbol, imin_, imax_, di_}]'
if isinstance(self, SympyFunction) and di.get_int_value() == 1:
whole_expr = Expression(self.get_name(), expr,
Expression('List', i, imin, imax))
sympy_expr = whole_expr.to_sympy()
# apply Together to produce results similar to Mathematica
result = sympy.together(sympy_expr)
result = from_sympy(result)
result = cancel(result)
if not result.same(whole_expr):
return result
index = imin.evaluate(evaluation)
imax = imax.evaluate(evaluation)
di = di.evaluate(evaluation)
result = []
while True:
cont = Expression('LessEqual', index, imax).evaluate(evaluation)
if cont == Symbol('False'):
break
if cont != Symbol('True'):
if self.throw_iterb:
evaluation.message(self.get_name(), 'iterb')
return
evaluation.check_stopped()
try:
item = dynamic_scoping(expr.evaluate, {i.name: index},
evaluation)
result.append(item)
except ContinueInterrupt:
if self.allow_loopcontrol:
pass
else:
raise
except BreakInterrupt:
if self.allow_loopcontrol:
break
else:
raise
index = Expression('Plus', index, di).evaluate(evaluation)
return self.get_result(result)
def apply(self, expr, evaluation):
'Factor[expr_]'
expr_sympy = expr.to_sympy()
if expr_sympy is None:
return None
try:
result = sympy.together(expr_sympy)
numer, denom = result.as_numer_denom()
if denom == 1:
result = sympy.factor(expr_sympy)
else:
result = sympy.factor(numer) / sympy.factor(denom)
except sympy.PolynomialError:
return expr
return from_sympy(result)
def apply(self, m, evaluation):
'Eigenvectors[m_]'
matrix = to_sympy_matrix(m)
if matrix is None or matrix.cols != matrix.rows or matrix.cols == 0:
return evaluation.message('Eigenvectors', 'matsq', m)
eigenvects = matrix.eigenvects()
eigenvects = sorted(eigenvects, key=lambda (val, c, vect): abs(val), reverse=True)
result = []
for val, count, basis in eigenvects:
vect = basis[0] # select first basis vector
vect = list(vect) # convert matrix to vector (list)
vect = from_sympy(vect)
result.append(vect)
result.extend([Expression('List', *([0] * matrix.rows))] * (matrix.rows - len(result)))
return Expression('List', *result)
def apply(self, m, b, evaluation):
'LeastSquares[m_, b_]'
matrix = to_sympy_matrix(m)
if matrix is None:
return evaluation.message('LeastSquares', 'matrix', m, 1)
b_vector = to_sympy_matrix(b)
if b_vector is None:
return evaluation.message('LeastSquares', 'matrix', b, 2)
try:
solution = matrix.solve_least_squares(b_vector) # default method = Cholesky
except NotImplementedError as e:
return evaluation.message('LeastSquares', 'underdetermined')
return from_sympy(solution)
def apply_iter(self, expr, i, imin, imax, di, evaluation):
'%(name)s[expr_, {i_Symbol, imin_, imax_, di_}]'
if di.get_int_value() == 1 and isinstance(self, SympyFunction):
whole_expr = Expression(self.get_name(), expr,
Expression('List', i, imin, imax))
sympy_expr = whole_expr.to_sympy()
# apply Together to produce results similar to Mathematica
result = sympy.together(sympy_expr)
result = from_sympy(result)
result = cancel(result)
if not result.same(whole_expr):
return result
index = imin.evaluate(evaluation).get_real_value()
imax = imax.evaluate(evaluation).get_real_value()
di = di.evaluate(evaluation).get_real_value()
if index is None or imax is None or di is None:
if self.throw_iterb:
evaluation.message(self.get_name(), 'iterb')
return
result = []
while index <= imax:
evaluation.check_stopped()
try:
item = dynamic_scoping(expr.evaluate,
{i.name: Number.from_mp(index)},
evaluation)
result.append(item)
except ContinueInterrupt:
if self.allow_loopcontrol:
pass
else:
raise
except BreakInterrupt:
if self.allow_loopcontrol:
break
else:
raise
index = index + di
return self.get_result(result)
def apply_dx(self, x, dx, evaluation):
"Rationalize[x_, dx_]"
py_x = x.to_sympy()
if py_x is None:
return x
py_dx = dx.to_sympy()
if (
py_dx is None
or (not py_dx.is_number)
or (not py_dx.is_real)
or py_dx.is_negative
):
return evaluation.message("Rationalize", "tolnn", dx)
elif py_dx == 0:
return from_sympy(self.find_exact(py_x))
a = self.approx_interval_continued_fraction(py_x - py_dx, py_x + py_dx)
sym_x = sympy.ntheory.continued_fraction_reduce(a)
return Rational(sym_x)
def apply(self, m, power, evaluation):
"MatrixPower[m_, power_]"
sympy_m = to_sympy_matrix(m)
if sympy_m is None:
return evaluation.message("MatrixPower", "matrix", m, 1)
sympy_power = power.to_sympy()
if sympy_power is None:
return
try:
res = sympy_m**sympy_power
except NotImplementedError:
return evaluation.message("MatrixPower",
"matrixpowernotimplemented", m)
except ValueError as e:
return evaluation.message("MatrixPower",
"matrixpowernotinvertible", m)
return from_sympy(res)
def _coefficient(name, expr, form, n, evaluation):
if expr == Symbol('Null') or form == Symbol('Null') or n == Symbol('Null'):
return Integer(0)
if not(isinstance(form, Symbol)) and not(isinstance(form, Expression)):
return evaluation.message(name, 'ivar', form)
sympy_exprs = expr.to_sympy().as_ordered_terms()
sympy_var = form.to_sympy()
sympy_n = n.to_sympy()
def combine_exprs(exprs):
result = 0
for e in exprs:
result += e
return result
# expand sub expressions if they contain variables
sympy_exprs = [sympy.expand(e) if sympy_var.free_symbols.issubset(e.free_symbols) else e for e in sympy_exprs]
sympy_expr = combine_exprs(sympy_exprs)
sympy_result = sympy_expr.coeff(sympy_var, sympy_n)
return from_sympy(sympy_result)
def apply(self, f, n, n0, evaluation, options={}):
'DiscreteLimit[f_, n_->n0_, OptionsPattern[DiscreteLimit]]'
f = f.to_sympy(convert_all_global_functions=True)
n = n.to_sympy()
n0 = n0.to_sympy()
if n0 != sympy.oo:
return
if f is None or n is None:
return
trials = options['System`Trials'].get_int_value()
if trials is None or trials <= 0:
evaluation.message('DiscreteLimit', 'dltrials')
trials = 5
try:
return from_sympy(sympy.limit_seq(f, n, trials))
except:
pass
def apply(self, m, b, evaluation):
'LinearSolve[m_, b_]'
matrix = matrix_data(m)
if matrix is None:
return
if not b.has_form('List', None):
return
if len(b.leaves) != len(matrix):
return evaluation.message('LinearSolve', 'lslc')
system = [m + [v] for m, v in zip(matrix, b.leaves)]
system = to_sympy_matrix(system)
if system is None:
return
syms = [sympy.Symbol('LinearSolve_var%d' % k, dummy=True) for k in range(system.cols - 1)]
sol = sympy.solve_linear_system(system, *syms)
if sol:
# substitute 0 for variables that are not in result dictionary
free_vars = dict((sym, sympy.Integer(0)) for sym in syms if sym not in sol)
sol.update(free_vars)
sol = [(sol[sym] if sym in free_vars else sol[sym].subs(free_vars)) for sym in syms]
return from_sympy(sol)
else:
return evaluation.message('LinearSolve', 'nosol')
def DixonPolynomial_apply(polys, xlist, alist):
xsym = [i.to_sympy() for i in xlist.leaves]
asym = [i.to_sympy() for i in alist.leaves]
res = DixonSub_apply(polys, xlist, alist)
if not isinstance(
res,
Expression): ####checking if the DixonSub_apply returned error
return res
rsym = res.leaves
dixon_list = [j.leaves for j in rsym
] ###two level Mathics list unpacked in 2 level python-list
dixon_sym = [[j.to_sympy() for j in i] for i in dixon_list]
dix_mat = sympy.Matrix(
dixon_sym) ###2-level python list to sympy.Matrix object
for i in range(1, len(xsym) + 1):
#####together has poor results in sympy so simplify was used
dix_mat[i, :] = sympy.simplify(
(dix_mat[i, :] - dix_mat[i - 1, :]) / (xsym[i - 1] - asym[i - 1]))
# print(rsym[0])
det_m = sympy.simplify(sympy.det(dix_mat))
return from_sympy(det_m)
def apply(self, m, l, evaluation):
'Norm[m_, l_]'
if isinstance(l, Symbol):
pass
elif isinstance(l, (Real, Integer)) and l.to_python() >= 1:
pass
else:
return evaluation.message('Norm', 'ptype', l)
l = l.to_sympy()
matrix = to_sympy_matrix(m)
if matrix is None:
return evaluation.message('Norm', 'nvm')
if len(matrix) == 0:
return
try:
res = matrix.norm(l)
except NotImplementedError:
return evaluation.message('Norm', 'normnotimplemented')
return from_sympy(res)
def apply(self, m, evaluation):
'PseudoInverse[m_]'
matrix = to_sympy_matrix(m)
pinv = matrix.pinv()
return from_sympy(pinv)
def apply(self, items, evaluation):
'Power[items__]'
items_sequence = items.get_sequence()
if len(items_sequence) == 2:
x, y = items_sequence
else:
return Expression('Power', *items_sequence)
if y.get_int_value() == 1:
return x
elif x.get_int_value() == 1:
return x
elif y.get_int_value() == 0:
if x.get_int_value() == 0:
evaluation.message('Power', 'indet', Expression('Power', x, y))
return Symbol('Indeterminate')
else:
return Integer(1)
elif x.has_form('Power', 2) and isinstance(y, Integer):
return Expression('Power', x.leaves[0],
Expression('Times', x.leaves[1], y))
elif x.has_form('Times', None) and isinstance(y, Integer):
return Expression(
'Times', *[Expression('Power', leaf, y) for leaf in x.leaves])
elif (isinstance(x, Number) and isinstance(y, Number)
and not (x.is_inexact() or y.is_inexact())):
sym_x, sym_y = x.to_sympy(), y.to_sympy()
try:
if sympy.re(sym_y) >= 0:
result = sym_x**sym_y
else:
if sym_x == 0:
evaluation.message('Power', 'infy')
return Symbol('ComplexInfinity')
result = sympy.Integer(1) / (sym_x**(-sym_y))
if isinstance(result, sympy.Pow):
result = result.simplify()
args = [from_sympy(expr) for expr in result.as_base_exp()]
result = Expression('Power', *args)
result = result.evaluate_leaves(evaluation)
return result
return from_sympy(result)
except ValueError:
return Expression('Power', x, y)
except ZeroDivisionError:
evaluation.message('Power', 'infy')
return Symbol('ComplexInfinity')
elif (isinstance(x, Number) and isinstance(y, Number)
and (x.is_inexact() or y.is_inexact())):
try:
prec = min_prec(x, y)
with mpmath.workprec(prec):
mp_x = sympy2mpmath(x.to_sympy())
mp_y = sympy2mpmath(y.to_sympy())
result = mp_x**mp_y
if isinstance(result, mpmath.mpf):
return Real(str(result), prec)
elif isinstance(result, mpmath.mpc):
return Complex(str(result.real), str(result.imag),
prec)
except ZeroDivisionError:
evaluation.message('Power', 'infy')
return Symbol('ComplexInfinity')
else:
numerified_items = items.numerify(evaluation)
return Expression('Power', *numerified_items.get_sequence())
def apply(self, items, evaluation):
'Times[items___]'
# TODO: Clean this up and optimise it
items = items.numerify(evaluation).get_sequence()
number = (sympy.Integer(1), sympy.Integer(0))
leaves = []
prec = min_prec(*items)
is_real = all([not isinstance(i, Complex) for i in items])
for item in items:
if isinstance(item, Number):
if isinstance(item, Complex):
sym_real, sym_imag = item.real.to_sympy(
), item.imag.to_sympy()
else:
sym_real, sym_imag = item.to_sympy(), sympy.Integer(0)
if prec is not None:
sym_real = sym_real.n(dps(prec))
sym_imag = sym_imag.n(dps(prec))
if sym_real.is_zero and sym_imag.is_zero and prec is None:
return Integer('0')
number = (number[0] * sym_real - number[1] * sym_imag,
number[0] * sym_imag + number[1] * sym_real)
elif leaves and item == leaves[-1]:
leaves[-1] = Expression('Power', leaves[-1], Integer(2))
elif (leaves and item.has_form('Power', 2)
and leaves[-1].has_form('Power', 2)
and item.leaves[0].same(leaves[-1].leaves[0])):
leaves[-1].leaves[1] = Expression('Plus', item.leaves[1],
leaves[-1].leaves[1])
elif (leaves and item.has_form('Power', 2)
and item.leaves[0].same(leaves[-1])):
leaves[-1] = Expression(
'Power', leaves[-1],
Expression('Plus', item.leaves[1], Integer(1)))
elif (leaves and leaves[-1].has_form('Power', 2)
and leaves[-1].leaves[0].same(item)):
leaves[-1] = Expression(
'Power', item,
Expression('Plus', Integer(1), leaves[-1].leaves[1]))
else:
leaves.append(item)
if number == (1, 0):
number = None
elif number == (-1, 0) and leaves and leaves[0].has_form('Plus', None):
leaves[0].leaves = [
Expression('Times', Integer(-1), leaf)
for leaf in leaves[0].leaves
]
number = None
if number is not None:
if number[1].is_zero and is_real:
leaves.insert(0, Number.from_mp(number[0], prec))
elif number[1].is_zero and number[1].is_Integer and prec is None:
leaves.insert(0, Number.from_mp(number[0], prec))
else:
leaves.insert(
0,
Complex(from_sympy(number[0]), from_sympy(number[1]),
prec))
if not leaves:
return Integer(1)
elif len(leaves) == 1:
return leaves[0]
else:
return Expression('Times', *leaves)
def apply_multiplevariable(self, f, vars, evaluation):
'Minimize[f_?NotListQ, vars_?ListQ]'
head_name = vars.get_head_name()
vars_or = vars
vars = vars.leaves
for var in vars:
if ((var.is_atom() and not var.is_symbol()) or # noqa
head_name in ('System`Plus', 'System`Times',
'System`Power') or 'System`Constant'
in var.get_attributes(evaluation.definitions)):
evaluation.message('Minimize', 'ivar', vars_or)
return
vars_sympy = [var.to_sympy() for var in vars]
sympy_f = f.to_sympy()
jacobian = [sympy.diff(sympy_f, x) for x in vars_sympy]
hessian = sympy.Matrix([[sympy.diff(deriv, x) for x in vars_sympy]
for deriv in jacobian])
candidates_tmp = sympy.solve(jacobian, vars_sympy, dict=True)
candidates = []
for candidate in candidates_tmp:
if len(candidate) != len(vars_sympy):
for variable in candidate:
for i in range(len(candidate), len(vars_sympy)):
candidate[variable] = candidate[variable].subs(
vars_sympy[i], 1)
for i in range(len(candidate), len(vars_sympy)):
candidate[vars_sympy[i]] = 1
candidates.append(candidate)
minimum_list = []
for candidate in candidates:
eigenvals = hessian.subs(candidate).eigenvals()
positives_eigenvalues = 0
negatives_eigenvalues = 0
for val in eigenvals:
if val.is_real:
if val < 0:
negatives_eigenvalues += 1
elif val >= 0:
positives_eigenvalues += 1
if positives_eigenvalues + negatives_eigenvalues != len(eigenvals):
continue
if positives_eigenvalues == len(eigenvals):
minimum_list.append(candidate)
return Expression(
'List',
*(Expression('List', from_sympy(
sympy_f.subs(minimum).simplify()), [
Expression('Rule', from_sympy(list(minimum.keys())[i]),
from_sympy(list(minimum.values())[i]))
for i in range(len(vars_sympy))
]) for minimum in minimum_list))
def apply_constraints(self, f, vars, evaluation):
'Minimize[f_?ListQ, vars_?ListQ]'
head_name = vars.get_head_name()
vars_or = vars
vars = vars.leaves
for var in vars:
if ((var.is_atom() and not var.is_symbol()) or # noqa
head_name in ('System`Plus', 'System`Times',
'System`Power') or 'System`Constant'
in var.get_attributes(evaluation.definitions)):
evaluation.message('Minimize', 'ivar', vars_or)
return
vars_sympy = [var.to_sympy() for var in vars]
constraints = [function for function in f.leaves]
objective_function = constraints[0].to_sympy()
constraints = constraints[1:]
g_functions = []
h_functions = []
g_variables = []
h_variables = []
for constraint in constraints:
left, right = constraint.leaves
head_name = constraint.get_head_name()
left = left.to_sympy()
right = right.to_sympy()
if head_name == 'System`LessEqual' or head_name == 'System`Less':
eq = left - right
eq = sympy.together(eq)
eq = sympy.cancel(eq)
g_functions.append(eq)
g_variables.append(
sympy.Symbol('kkt_g' + str(len(g_variables))))
elif head_name == 'System`GreaterEqual' or head_name == 'System`Greater':
eq = -1 * (left - right)
eq = sympy.together(eq)
eq = sympy.cancel(eq)
g_functions.append(eq)
g_variables.append(
sympy.Symbol('kkt_g' + str(len(g_variables))))
elif head_name == 'System`Equal':
eq = left - right
eq = sympy.together(eq)
eq = sympy.cancel(eq)
h_functions.append(eq)
h_variables.append(
sympy.Symbol('kkt_h' + str(len(h_variables))))
equations = []
for variable in vars_sympy:
equation = sympy.diff(objective_function, variable)
for i in range(len(g_variables)):
g_variable = g_variables[i]
g_function = g_functions[i]
equation = equation + g_variable * sympy.diff(
g_function, variable)
for i in range(len(h_variables)):
h_variable = h_variables[i]
h_function = h_functions[i]
equation = equation + h_variable * sympy.diff(
h_function, variable)
equations.append(equation)
for i in range(len(g_variables)):
g_variable = g_variables[i]
g_function = g_functions[i]
equations.append(g_variable * g_function)
for i in range(len(h_variables)):
h_variable = h_variables[i]
h_function = h_functions[i]
equations.append(h_variable * h_function)
all_variables = vars_sympy + g_variables + h_variables
candidates_tmp = sympy.solve(equations, all_variables, dict=True)
candidates = []
for candidate in candidates_tmp:
if len(candidate) != len(vars_sympy):
for variable in candidate:
for i in range(len(candidate), len(vars_sympy)):
candidate[variable] = candidate[variable].subs(
vars_sympy[i], 1)
for i in range(len(candidate), len(vars_sympy)):
candidate[vars_sympy[i]] = 1
candidates.append(candidate)
kkt_candidates = []
for candidate in candidates:
kkt_ok = True
sum_constraints = 0
for i in range(len(g_variables)):
g_variable = g_variables[i]
g_function = g_functions[i]
if candidate[g_variable] < 0:
kkt_ok = False
if candidate[g_variable] * g_function.subs(candidate) != 0:
kkt_ok = False
sum_constraints = sum_constraints + candidate[g_variable]
for i in range(len(h_variables)):
h_variable = h_variables[i]
h_function = h_functions[i]
sum_constraints = sum_constraints + abs(candidate[h_variable])
if sum_constraints <= 0:
kkt_ok = False
if not kkt_ok:
continue
kkt_candidates.append(candidate)
hessian = sympy.Matrix([[sympy.diff(deriv, x) for x in all_variables]
for deriv in equations])
for i in range(0, len(all_variables) - len(vars_sympy)):
hessian.col_del(len(all_variables) - i - 1)
hessian.row_del(len(all_variables) - i - 1)
minimum_list = []
for candidate in kkt_candidates:
eigenvals = hessian.subs(candidate).eigenvals()
positives_eigenvalues = 0
negatives_eigenvalues = 0
for val in eigenvals:
val = complex(sympy.N(val, chop=True))
if val.imag == 0:
val = val.real
if val < 0:
negatives_eigenvalues += 1
elif val > 0:
positives_eigenvalues += 1
if positives_eigenvalues + negatives_eigenvalues != len(eigenvals):
continue
if positives_eigenvalues == len(eigenvals):
for g_variable in g_variables:
del candidate[g_variable]
for h_variable in h_variables:
del candidate[h_variable]
minimum_list.append(candidate)
return Expression(
'List',
*(Expression(
'List',
from_sympy(objective_function.subs(minimum).simplify()), [
Expression('Rule', from_sympy(list(minimum.keys())[i]),
from_sympy(list(minimum.values())[i]))
for i in range(len(vars_sympy))
]) for minimum in minimum_list))
def apply(self, eqns, a, n, evaluation):
'RSolve[eqns_, a_, n_]'
# TODO: Do this with rules?
if not eqns.has_form('List', None):
eqns = Expression('List', eqns)
if len(eqns.leaves) == 0:
return
for eqn in eqns.leaves:
if eqn.get_head_name() != 'System`Equal':
evaluation.message('RSolve', 'deqn', eqn)
return
if (n.is_atom() and not n.is_symbol()) or \
n.get_head_name() in ('System`Plus', 'System`Times',
'System`Power') or \
'System`Constant' in n.get_attributes(evaluation.definitions):
# TODO: Factor out this check for dsvar into a separate
# function. DSolve uses this too.
evaluation.message('RSolve', 'dsvar')
return
try:
a.leaves
function_form = None
func = a
except AttributeError:
func = Expression(a, n)
function_form = Expression('List', n)
if func.is_atom() or len(func.leaves) != 1:
evaluation.message('RSolve', 'dsfun', a)
if n not in func.leaves:
evaluation.message('DSolve', 'deqx')
# Seperate relations from conditions
conditions = {}
def is_relation(eqn):
left, right = eqn.leaves
for l, r in [(left, right), (right, left)]:
if (left.get_head_name() == func.get_head_name() and # noqa
len(left.leaves) == 1 and isinstance(
l.leaves[0].to_python(), int) and r.is_numeric()):
r_sympy = r.to_sympy()
if r_sympy is None:
raise ValueError
conditions[l.leaves[0].to_python()] = r_sympy
return False
return True
# evaluate is_relation on all leaves to store conditions
try:
relations = [leaf for leaf in eqns.leaves if is_relation(leaf)]
except ValueError:
return
relation = relations[0]
left, right = relation.leaves
relation = Expression('Plus', left,
Expression('Times', -1,
right)).evaluate(evaluation)
sym_eq = relation.to_sympy(
converted_functions=set([func.get_head_name()]))
if sym_eq is None:
return
sym_n = sympy.core.symbols(str(sympy_symbol_prefix + n.name))
sym_func = sympy.Function(
str(sympy_symbol_prefix + func.get_head_name()))(sym_n)
sym_conds = {}
for cond in conditions:
sym_conds[sympy.Function(str(
sympy_symbol_prefix + func.get_head_name()))(cond)] = \
conditions[cond]
try:
# Sympy raises error when given empty conditions. Fixed in
# upcomming sympy release.
if sym_conds != {}:
sym_result = sympy.rsolve(sym_eq, sym_func, sym_conds)
else:
sym_result = sympy.rsolve(sym_eq, sym_func)
if not isinstance(sym_result, list):
sym_result = [sym_result]
except ValueError:
return
if function_form is None:
return Expression(
'List', *[
Expression('List', Expression('Rule', a, from_sympy(soln)))
for soln in sym_result
])
else:
return Expression(
'List', *[
Expression(
'List',
Expression(
'Rule', a,
Expression('Function', function_form,
from_sympy(soln))))
for soln in sym_result
])
def apply(self, x, evaluation):
'Ceiling[x_]'
x = x.to_sympy()
return from_sympy(sympy.ceiling(x))
