You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

1870 lines
51 KiB

"""Formal Power Series"""
from collections import defaultdict
from sympy.core.numbers import (nan, oo, zoo)
from sympy.core.add import Add
from sympy.core.expr import Expr
from sympy.core.function import Derivative, Function, expand
from sympy.core.mul import Mul
from sympy.core.numbers import Rational
from sympy.core.relational import Eq
from sympy.sets.sets import Interval
from sympy.core.singleton import S
from sympy.core.symbol import Wild, Dummy, symbols, Symbol
from sympy.core.sympify import sympify
from sympy.discrete.convolutions import convolution
from sympy.functions.combinatorial.factorials import binomial, factorial, rf
from sympy.functions.combinatorial.numbers import bell
from sympy.functions.elementary.integers import floor, frac, ceiling
from sympy.functions.elementary.miscellaneous import Min, Max
from sympy.functions.elementary.piecewise import Piecewise
from sympy.series.limits import Limit
from sympy.series.order import Order
from sympy.series.sequences import sequence
from sympy.series.series_class import SeriesBase
from sympy.utilities.iterables import iterable
def rational_algorithm(f, x, k, order=4, full=False):
"""
Rational algorithm for computing
formula of coefficients of Formal Power Series
of a function.
Explanation
===========
Applicable when f(x) or some derivative of f(x)
is a rational function in x.
:func:`rational_algorithm` uses :func:`~.apart` function for partial fraction
decomposition. :func:`~.apart` by default uses 'undetermined coefficients
method'. By setting ``full=True``, 'Bronstein's algorithm' can be used
instead.
Looks for derivative of a function up to 4'th order (by default).
This can be overridden using order option.
Parameters
==========
x : Symbol
order : int, optional
Order of the derivative of ``f``, Default is 4.
full : bool
Returns
=======
formula : Expr
ind : Expr
Independent terms.
order : int
full : bool
Examples
========
>>> from sympy import log, atan
>>> from sympy.series.formal import rational_algorithm as ra
>>> from sympy.abc import x, k
>>> ra(1 / (1 - x), x, k)
(1, 0, 0)
>>> ra(log(1 + x), x, k)
(-1/((-1)**k*k), 0, 1)
>>> ra(atan(x), x, k, full=True)
((-I/(2*(-I)**k) + I/(2*I**k))/k, 0, 1)
Notes
=====
By setting ``full=True``, range of admissible functions to be solved using
``rational_algorithm`` can be increased. This option should be used
carefully as it can significantly slow down the computation as ``doit`` is
performed on the :class:`~.RootSum` object returned by the :func:`~.apart`
function. Use ``full=False`` whenever possible.
See Also
========
sympy.polys.partfrac.apart
References
==========
.. [1] Formal Power Series - Dominik Gruntz, Wolfram Koepf
.. [2] Power Series in Computer Algebra - Wolfram Koepf
"""
from sympy.polys import RootSum, apart
from sympy.integrals import integrate
diff = f
ds = [] # list of diff
for i in range(order + 1):
if i:
diff = diff.diff(x)
if diff.is_rational_function(x):
coeff, sep = S.Zero, S.Zero
terms = apart(diff, x, full=full)
if terms.has(RootSum):
terms = terms.doit()
for t in Add.make_args(terms):
num, den = t.as_numer_denom()
if not den.has(x):
sep += t
else:
if isinstance(den, Mul):
# m*(n*x - a)**j -> (n*x - a)**j
ind = den.as_independent(x)
den = ind[1]
num /= ind[0]
# (n*x - a)**j -> (x - b)
den, j = den.as_base_exp()
a, xterm = den.as_coeff_add(x)
# term -> m/x**n
if not a:
sep += t
continue
xc = xterm[0].coeff(x)
a /= -xc
num /= xc**j
ak = ((-1)**j * num *
binomial(j + k - 1, k).rewrite(factorial) /
a**(j + k))
coeff += ak
# Hacky, better way?
if coeff.is_zero:
return None
if (coeff.has(x) or coeff.has(zoo) or coeff.has(oo) or
coeff.has(nan)):
return None
for j in range(i):
coeff = (coeff / (k + j + 1))
sep = integrate(sep, x)
sep += (ds.pop() - sep).limit(x, 0) # constant of integration
return (coeff.subs(k, k - i), sep, i)
else:
ds.append(diff)
return None
def rational_independent(terms, x):
"""
Returns a list of all the rationally independent terms.
Examples
========
>>> from sympy import sin, cos
>>> from sympy.series.formal import rational_independent
>>> from sympy.abc import x
>>> rational_independent([cos(x), sin(x)], x)
[cos(x), sin(x)]
>>> rational_independent([x**2, sin(x), x*sin(x), x**3], x)
[x**3 + x**2, x*sin(x) + sin(x)]
"""
if not terms:
return []
ind = terms[0:1]
for t in terms[1:]:
n = t.as_independent(x)[1]
for i, term in enumerate(ind):
d = term.as_independent(x)[1]
q = (n / d).cancel()
if q.is_rational_function(x):
ind[i] += t
break
else:
ind.append(t)
return ind
def simpleDE(f, x, g, order=4):
r"""
Generates simple DE.
Explanation
===========
DE is of the form
.. math::
f^k(x) + \sum\limits_{j=0}^{k-1} A_j f^j(x) = 0
where :math:`A_j` should be rational function in x.
Generates DE's upto order 4 (default). DE's can also have free parameters.
By increasing order, higher order DE's can be found.
Yields a tuple of (DE, order).
"""
from sympy.solvers.solveset import linsolve
a = symbols('a:%d' % (order))
def _makeDE(k):
eq = f.diff(x, k) + Add(*[a[i]*f.diff(x, i) for i in range(0, k)])
DE = g(x).diff(x, k) + Add(*[a[i]*g(x).diff(x, i) for i in range(0, k)])
return eq, DE
found = False
for k in range(1, order + 1):
eq, DE = _makeDE(k)
eq = eq.expand()
terms = eq.as_ordered_terms()
ind = rational_independent(terms, x)
if found or len(ind) == k:
sol = dict(zip(a, (i for s in linsolve(ind, a[:k]) for i in s)))
if sol:
found = True
DE = DE.subs(sol)
DE = DE.as_numer_denom()[0]
DE = DE.factor().as_coeff_mul(Derivative)[1][0]
yield DE.collect(Derivative(g(x))), k
def exp_re(DE, r, k):
"""Converts a DE with constant coefficients (explike) into a RE.
Explanation
===========
Performs the substitution:
.. math::
f^j(x) \\to r(k + j)
Normalises the terms so that lowest order of a term is always r(k).
Examples
========
>>> from sympy import Function, Derivative
>>> from sympy.series.formal import exp_re
>>> from sympy.abc import x, k
>>> f, r = Function('f'), Function('r')
>>> exp_re(-f(x) + Derivative(f(x)), r, k)
-r(k) + r(k + 1)
>>> exp_re(Derivative(f(x), x) + Derivative(f(x), (x, 2)), r, k)
r(k) + r(k + 1)
See Also
========
sympy.series.formal.hyper_re
"""
RE = S.Zero
g = DE.atoms(Function).pop()
mini = None
for t in Add.make_args(DE):
coeff, d = t.as_independent(g)
if isinstance(d, Derivative):
j = d.derivative_count
else:
j = 0
if mini is None or j < mini:
mini = j
RE += coeff * r(k + j)
if mini:
RE = RE.subs(k, k - mini)
return RE
def hyper_re(DE, r, k):
"""
Converts a DE into a RE.
Explanation
===========
Performs the substitution:
.. math::
x^l f^j(x) \\to (k + 1 - l)_j . a_{k + j - l}
Normalises the terms so that lowest order of a term is always r(k).
Examples
========
>>> from sympy import Function, Derivative
>>> from sympy.series.formal import hyper_re
>>> from sympy.abc import x, k
>>> f, r = Function('f'), Function('r')
>>> hyper_re(-f(x) + Derivative(f(x)), r, k)
(k + 1)*r(k + 1) - r(k)
>>> hyper_re(-x*f(x) + Derivative(f(x), (x, 2)), r, k)
(k + 2)*(k + 3)*r(k + 3) - r(k)
See Also
========
sympy.series.formal.exp_re
"""
RE = S.Zero
g = DE.atoms(Function).pop()
x = g.atoms(Symbol).pop()
mini = None
for t in Add.make_args(DE.expand()):
coeff, d = t.as_independent(g)
c, v = coeff.as_independent(x)
l = v.as_coeff_exponent(x)[1]
if isinstance(d, Derivative):
j = d.derivative_count
else:
j = 0
RE += c * rf(k + 1 - l, j) * r(k + j - l)
if mini is None or j - l < mini:
mini = j - l
RE = RE.subs(k, k - mini)
m = Wild('m')
return RE.collect(r(k + m))
def _transformation_a(f, x, P, Q, k, m, shift):
f *= x**(-shift)
P = P.subs(k, k + shift)
Q = Q.subs(k, k + shift)
return f, P, Q, m
def _transformation_c(f, x, P, Q, k, m, scale):
f = f.subs(x, x**scale)
P = P.subs(k, k / scale)
Q = Q.subs(k, k / scale)
m *= scale
return f, P, Q, m
def _transformation_e(f, x, P, Q, k, m):
f = f.diff(x)
P = P.subs(k, k + 1) * (k + m + 1)
Q = Q.subs(k, k + 1) * (k + 1)
return f, P, Q, m
def _apply_shift(sol, shift):
return [(res, cond + shift) for res, cond in sol]
def _apply_scale(sol, scale):
return [(res, cond / scale) for res, cond in sol]
def _apply_integrate(sol, x, k):
return [(res / ((cond + 1)*(cond.as_coeff_Add()[1].coeff(k))), cond + 1)
for res, cond in sol]
def _compute_formula(f, x, P, Q, k, m, k_max):
"""Computes the formula for f."""
from sympy.polys import roots
sol = []
for i in range(k_max + 1, k_max + m + 1):
if (i < 0) == True:
continue
r = f.diff(x, i).limit(x, 0) / factorial(i)
if r.is_zero:
continue
kterm = m*k + i
res = r
p = P.subs(k, kterm)
q = Q.subs(k, kterm)
c1 = p.subs(k, 1/k).leadterm(k)[0]
c2 = q.subs(k, 1/k).leadterm(k)[0]
res *= (-c1 / c2)**k
res *= Mul(*[rf(-r, k)**mul for r, mul in roots(p, k).items()])
res /= Mul(*[rf(-r, k)**mul for r, mul in roots(q, k).items()])
sol.append((res, kterm))
return sol
def _rsolve_hypergeometric(f, x, P, Q, k, m):
"""
Recursive wrapper to rsolve_hypergeometric.
Explanation
===========
Returns a Tuple of (formula, series independent terms,
maximum power of x in independent terms) if successful
otherwise ``None``.
See :func:`rsolve_hypergeometric` for details.
"""
from sympy.polys import lcm, roots
from sympy.integrals import integrate
# transformation - c
proots, qroots = roots(P, k), roots(Q, k)
all_roots = dict(proots)
all_roots.update(qroots)
scale = lcm([r.as_numer_denom()[1] for r, t in all_roots.items()
if r.is_rational])
f, P, Q, m = _transformation_c(f, x, P, Q, k, m, scale)
# transformation - a
qroots = roots(Q, k)
if qroots:
k_min = Min(*qroots.keys())
else:
k_min = S.Zero
shift = k_min + m
f, P, Q, m = _transformation_a(f, x, P, Q, k, m, shift)
l = (x*f).limit(x, 0)
if not isinstance(l, Limit) and l != 0: # Ideally should only be l != 0
return None
qroots = roots(Q, k)
if qroots:
k_max = Max(*qroots.keys())
else:
k_max = S.Zero
ind, mp = S.Zero, -oo
for i in range(k_max + m + 1):
r = f.diff(x, i).limit(x, 0) / factorial(i)
if r.is_finite is False:
old_f = f
f, P, Q, m = _transformation_a(f, x, P, Q, k, m, i)
f, P, Q, m = _transformation_e(f, x, P, Q, k, m)
sol, ind, mp = _rsolve_hypergeometric(f, x, P, Q, k, m)
sol = _apply_integrate(sol, x, k)
sol = _apply_shift(sol, i)
ind = integrate(ind, x)
ind += (old_f - ind).limit(x, 0) # constant of integration
mp += 1
return sol, ind, mp
elif r:
ind += r*x**(i + shift)
pow_x = Rational((i + shift), scale)
if pow_x > mp:
mp = pow_x # maximum power of x
ind = ind.subs(x, x**(1/scale))
sol = _compute_formula(f, x, P, Q, k, m, k_max)
sol = _apply_shift(sol, shift)
sol = _apply_scale(sol, scale)
return sol, ind, mp
def rsolve_hypergeometric(f, x, P, Q, k, m):
"""
Solves RE of hypergeometric type.
Explanation
===========
Attempts to solve RE of the form
Q(k)*a(k + m) - P(k)*a(k)
Transformations that preserve Hypergeometric type:
a. x**n*f(x): b(k + m) = R(k - n)*b(k)
b. f(A*x): b(k + m) = A**m*R(k)*b(k)
c. f(x**n): b(k + n*m) = R(k/n)*b(k)
d. f(x**(1/m)): b(k + 1) = R(k*m)*b(k)
e. f'(x): b(k + m) = ((k + m + 1)/(k + 1))*R(k + 1)*b(k)
Some of these transformations have been used to solve the RE.
Returns
=======
formula : Expr
ind : Expr
Independent terms.
order : int
Examples
========
>>> from sympy import exp, ln, S
>>> from sympy.series.formal import rsolve_hypergeometric as rh
>>> from sympy.abc import x, k
>>> rh(exp(x), x, -S.One, (k + 1), k, 1)
(Piecewise((1/factorial(k), Eq(Mod(k, 1), 0)), (0, True)), 1, 1)
>>> rh(ln(1 + x), x, k**2, k*(k + 1), k, 1)
(Piecewise(((-1)**(k - 1)*factorial(k - 1)/RisingFactorial(2, k - 1),
Eq(Mod(k, 1), 0)), (0, True)), x, 2)
References
==========
.. [1] Formal Power Series - Dominik Gruntz, Wolfram Koepf
.. [2] Power Series in Computer Algebra - Wolfram Koepf
"""
result = _rsolve_hypergeometric(f, x, P, Q, k, m)
if result is None:
return None
sol_list, ind, mp = result
sol_dict = defaultdict(lambda: S.Zero)
for res, cond in sol_list:
j, mk = cond.as_coeff_Add()
c = mk.coeff(k)
if j.is_integer is False:
res *= x**frac(j)
j = floor(j)
res = res.subs(k, (k - j) / c)
cond = Eq(k % c, j % c)
sol_dict[cond] += res # Group together formula for same conditions
sol = []
for cond, res in sol_dict.items():
sol.append((res, cond))
sol.append((S.Zero, True))
sol = Piecewise(*sol)
if mp is -oo:
s = S.Zero
elif mp.is_integer is False:
s = ceiling(mp)
else:
s = mp + 1
# save all the terms of
# form 1/x**k in ind
if s < 0:
ind += sum(sequence(sol * x**k, (k, s, -1)))
s = S.Zero
return (sol, ind, s)
def _solve_hyper_RE(f, x, RE, g, k):
"""See docstring of :func:`rsolve_hypergeometric` for details."""
terms = Add.make_args(RE)
if len(terms) == 2:
gs = list(RE.atoms(Function))
P, Q = map(RE.coeff, gs)
m = gs[1].args[0] - gs[0].args[0]
if m < 0:
P, Q = Q, P
m = abs(m)
return rsolve_hypergeometric(f, x, P, Q, k, m)
def _solve_explike_DE(f, x, DE, g, k):
"""Solves DE with constant coefficients."""
from sympy.solvers import rsolve
for t in Add.make_args(DE):
coeff, d = t.as_independent(g)
if coeff.free_symbols:
return
RE = exp_re(DE, g, k)
init = {}
for i in range(len(Add.make_args(RE))):
if i:
f = f.diff(x)
init[g(k).subs(k, i)] = f.limit(x, 0)
sol = rsolve(RE, g(k), init)
if sol:
return (sol / factorial(k), S.Zero, S.Zero)
def _solve_simple(f, x, DE, g, k):
"""Converts DE into RE and solves using :func:`rsolve`."""
from sympy.solvers import rsolve
RE = hyper_re(DE, g, k)
init = {}
for i in range(len(Add.make_args(RE))):
if i:
f = f.diff(x)
init[g(k).subs(k, i)] = f.limit(x, 0) / factorial(i)
sol = rsolve(RE, g(k), init)
if sol:
return (sol, S.Zero, S.Zero)
def _transform_explike_DE(DE, g, x, order, syms):
"""Converts DE with free parameters into DE with constant coefficients."""
from sympy.solvers.solveset import linsolve
eq = []
highest_coeff = DE.coeff(Derivative(g(x), x, order))
for i in range(order):
coeff = DE.coeff(Derivative(g(x), x, i))
coeff = (coeff / highest_coeff).expand().collect(x)
for t in Add.make_args(coeff):
eq.append(t)
temp = []
for e in eq:
if e.has(x):
break
elif e.has(Symbol):
temp.append(e)
else:
eq = temp
if eq:
sol = dict(zip(syms, (i for s in linsolve(eq, list(syms)) for i in s)))
if sol:
DE = DE.subs(sol)
DE = DE.factor().as_coeff_mul(Derivative)[1][0]
DE = DE.collect(Derivative(g(x)))
return DE
def _transform_DE_RE(DE, g, k, order, syms):
"""Converts DE with free parameters into RE of hypergeometric type."""
from sympy.solvers.solveset import linsolve
RE = hyper_re(DE, g, k)
eq = []
for i in range(1, order):
coeff = RE.coeff(g(k + i))
eq.append(coeff)
sol = dict(zip(syms, (i for s in linsolve(eq, list(syms)) for i in s)))
if sol:
m = Wild('m')
RE = RE.subs(sol)
RE = RE.factor().as_numer_denom()[0].collect(g(k + m))
RE = RE.as_coeff_mul(g)[1][0]
for i in range(order): # smallest order should be g(k)
if RE.coeff(g(k + i)) and i:
RE = RE.subs(k, k - i)
break
return RE
def solve_de(f, x, DE, order, g, k):
"""
Solves the DE.
Explanation
===========
Tries to solve DE by either converting into a RE containing two terms or
converting into a DE having constant coefficients.
Returns
=======
formula : Expr
ind : Expr
Independent terms.
order : int
Examples
========
>>> from sympy import Derivative as D, Function
>>> from sympy import exp, ln
>>> from sympy.series.formal import solve_de
>>> from sympy.abc import x, k
>>> f = Function('f')
>>> solve_de(exp(x), x, D(f(x), x) - f(x), 1, f, k)
(Piecewise((1/factorial(k), Eq(Mod(k, 1), 0)), (0, True)), 1, 1)
>>> solve_de(ln(1 + x), x, (x + 1)*D(f(x), x, 2) + D(f(x)), 2, f, k)
(Piecewise(((-1)**(k - 1)*factorial(k - 1)/RisingFactorial(2, k - 1),
Eq(Mod(k, 1), 0)), (0, True)), x, 2)
"""
sol = None
syms = DE.free_symbols.difference({g, x})
if syms:
RE = _transform_DE_RE(DE, g, k, order, syms)
else:
RE = hyper_re(DE, g, k)
if not RE.free_symbols.difference({k}):
sol = _solve_hyper_RE(f, x, RE, g, k)
if sol:
return sol
if syms:
DE = _transform_explike_DE(DE, g, x, order, syms)
if not DE.free_symbols.difference({x}):
sol = _solve_explike_DE(f, x, DE, g, k)
if sol:
return sol
def hyper_algorithm(f, x, k, order=4):
"""
Hypergeometric algorithm for computing Formal Power Series.
Explanation
===========
Steps:
* Generates DE
* Convert the DE into RE
* Solves the RE
Examples
========
>>> from sympy import exp, ln
>>> from sympy.series.formal import hyper_algorithm
>>> from sympy.abc import x, k
>>> hyper_algorithm(exp(x), x, k)
(Piecewise((1/factorial(k), Eq(Mod(k, 1), 0)), (0, True)), 1, 1)
>>> hyper_algorithm(ln(1 + x), x, k)
(Piecewise(((-1)**(k - 1)*factorial(k - 1)/RisingFactorial(2, k - 1),
Eq(Mod(k, 1), 0)), (0, True)), x, 2)
See Also
========
sympy.series.formal.simpleDE
sympy.series.formal.solve_de
"""
g = Function('g')
des = [] # list of DE's
sol = None
for DE, i in simpleDE(f, x, g, order):
if DE is not None:
sol = solve_de(f, x, DE, i, g, k)
if sol:
return sol
if not DE.free_symbols.difference({x}):
des.append(DE)
# If nothing works
# Try plain rsolve
for DE in des:
sol = _solve_simple(f, x, DE, g, k)
if sol:
return sol
def _compute_fps(f, x, x0, dir, hyper, order, rational, full):
"""Recursive wrapper to compute fps.
See :func:`compute_fps` for details.
"""
if x0 in [S.Infinity, S.NegativeInfinity]:
dir = S.One if x0 is S.Infinity else -S.One
temp = f.subs(x, 1/x)
result = _compute_fps(temp, x, 0, dir, hyper, order, rational, full)
if result is None:
return None
return (result[0], result[1].subs(x, 1/x), result[2].subs(x, 1/x))
elif x0 or dir == -S.One:
if dir == -S.One:
rep = -x + x0
rep2 = -x
rep2b = x0
else:
rep = x + x0
rep2 = x
rep2b = -x0
temp = f.subs(x, rep)
result = _compute_fps(temp, x, 0, S.One, hyper, order, rational, full)
if result is None:
return None
return (result[0], result[1].subs(x, rep2 + rep2b),
result[2].subs(x, rep2 + rep2b))
if f.is_polynomial(x):
k = Dummy('k')
ak = sequence(Coeff(f, x, k), (k, 1, oo))
xk = sequence(x**k, (k, 0, oo))
ind = f.coeff(x, 0)
return ak, xk, ind
# Break instances of Add
# this allows application of different
# algorithms on different terms increasing the
# range of admissible functions.
if isinstance(f, Add):
result = False
ak = sequence(S.Zero, (0, oo))
ind, xk = S.Zero, None
for t in Add.make_args(f):
res = _compute_fps(t, x, 0, S.One, hyper, order, rational, full)
if res:
if not result:
result = True
xk = res[1]
if res[0].start > ak.start:
seq = ak
s, f = ak.start, res[0].start
else:
seq = res[0]
s, f = res[0].start, ak.start
save = Add(*[z[0]*z[1] for z in zip(seq[0:(f - s)], xk[s:f])])
ak += res[0]
ind += res[2] + save
else:
ind += t
if result:
return ak, xk, ind
return None
# The symbolic term - symb, if present, is being separated from the function
# Otherwise symb is being set to S.One
syms = f.free_symbols.difference({x})
(f, symb) = expand(f).as_independent(*syms)
result = None
# from here on it's x0=0 and dir=1 handling
k = Dummy('k')
if rational:
result = rational_algorithm(f, x, k, order, full)
if result is None and hyper:
result = hyper_algorithm(f, x, k, order)
if result is None:
return None
from sympy.simplify.powsimp import powsimp
if symb.is_zero:
symb = S.One
else:
symb = powsimp(symb)
ak = sequence(result[0], (k, result[2], oo))
xk_formula = powsimp(x**k * symb)
xk = sequence(xk_formula, (k, 0, oo))
ind = powsimp(result[1] * symb)
return ak, xk, ind
def compute_fps(f, x, x0=0, dir=1, hyper=True, order=4, rational=True,
full=False):
"""
Computes the formula for Formal Power Series of a function.
Explanation
===========
Tries to compute the formula by applying the following techniques
(in order):
* rational_algorithm
* Hypergeometric algorithm
Parameters
==========
x : Symbol
x0 : number, optional
Point to perform series expansion about. Default is 0.
dir : {1, -1, '+', '-'}, optional
If dir is 1 or '+' the series is calculated from the right and
for -1 or '-' the series is calculated from the left. For smooth
functions this flag will not alter the results. Default is 1.
hyper : {True, False}, optional
Set hyper to False to skip the hypergeometric algorithm.
By default it is set to False.
order : int, optional
Order of the derivative of ``f``, Default is 4.
rational : {True, False}, optional
Set rational to False to skip rational algorithm. By default it is set
to True.
full : {True, False}, optional
Set full to True to increase the range of rational algorithm.
See :func:`rational_algorithm` for details. By default it is set to
False.
Returns
=======
ak : sequence
Sequence of coefficients.
xk : sequence
Sequence of powers of x.
ind : Expr
Independent terms.
mul : Pow
Common terms.
See Also
========
sympy.series.formal.rational_algorithm
sympy.series.formal.hyper_algorithm
"""
f = sympify(f)
x = sympify(x)
if not f.has(x):
return None
x0 = sympify(x0)
if dir == '+':
dir = S.One
elif dir == '-':
dir = -S.One
elif dir not in [S.One, -S.One]:
raise ValueError("Dir must be '+' or '-'")
else:
dir = sympify(dir)
return _compute_fps(f, x, x0, dir, hyper, order, rational, full)
class Coeff(Function):
"""
Coeff(p, x, n) represents the nth coefficient of the polynomial p in x
"""
@classmethod
def eval(cls, p, x, n):
if p.is_polynomial(x) and n.is_integer:
return p.coeff(x, n)
class FormalPowerSeries(SeriesBase):
"""
Represents Formal Power Series of a function.
Explanation
===========
No computation is performed. This class should only to be used to represent
a series. No checks are performed.
For computing a series use :func:`fps`.
See Also
========
sympy.series.formal.fps
"""
def __new__(cls, *args):
args = map(sympify, args)
return Expr.__new__(cls, *args)
def __init__(self, *args):
ak = args[4][0]
k = ak.variables[0]
self.ak_seq = sequence(ak.formula, (k, 1, oo))
self.fact_seq = sequence(factorial(k), (k, 1, oo))
self.bell_coeff_seq = self.ak_seq * self.fact_seq
self.sign_seq = sequence((-1, 1), (k, 1, oo))
@property
def function(self):
return self.args[0]
@property
def x(self):
return self.args[1]
@property
def x0(self):
return self.args[2]
@property
def dir(self):
return self.args[3]
@property
def ak(self):
return self.args[4][0]
@property
def xk(self):
return self.args[4][1]
@property
def ind(self):
return self.args[4][2]
@property
def interval(self):
return Interval(0, oo)
@property
def start(self):
return self.interval.inf
@property
def stop(self):
return self.interval.sup
@property
def length(self):
return oo
@property
def infinite(self):
"""Returns an infinite representation of the series"""
from sympy.concrete import Sum
ak, xk = self.ak, self.xk
k = ak.variables[0]
inf_sum = Sum(ak.formula * xk.formula, (k, ak.start, ak.stop))
return self.ind + inf_sum
def _get_pow_x(self, term):
"""Returns the power of x in a term."""
xterm, pow_x = term.as_independent(self.x)[1].as_base_exp()
if not xterm.has(self.x):
return S.Zero
return pow_x
def polynomial(self, n=6):
"""
Truncated series as polynomial.
Explanation
===========
Returns series expansion of ``f`` upto order ``O(x**n)``
as a polynomial(without ``O`` term).
"""
terms = []
sym = self.free_symbols
for i, t in enumerate(self):
xp = self._get_pow_x(t)
if xp.has(*sym):
xp = xp.as_coeff_add(*sym)[0]
if xp >= n:
break
elif xp.is_integer is True and i == n + 1:
break
elif t is not S.Zero:
terms.append(t)
return Add(*terms)
def truncate(self, n=6):
"""
Truncated series.
Explanation
===========
Returns truncated series expansion of f upto
order ``O(x**n)``.
If n is ``None``, returns an infinite iterator.
"""
if n is None:
return iter(self)
x, x0 = self.x, self.x0
pt_xk = self.xk.coeff(n)
if x0 is S.NegativeInfinity:
x0 = S.Infinity
return self.polynomial(n) + Order(pt_xk, (x, x0))
def zero_coeff(self):
return self._eval_term(0)
def _eval_term(self, pt):
try:
pt_xk = self.xk.coeff(pt)
pt_ak = self.ak.coeff(pt).simplify() # Simplify the coefficients
except IndexError:
term = S.Zero
else:
term = (pt_ak * pt_xk)
if self.ind:
ind = S.Zero
sym = self.free_symbols
for t in Add.make_args(self.ind):
pow_x = self._get_pow_x(t)
if pow_x.has(*sym):
pow_x = pow_x.as_coeff_add(*sym)[0]
if pt == 0 and pow_x < 1:
ind += t
elif pow_x >= pt and pow_x < pt + 1:
ind += t
term += ind
return term.collect(self.x)
def _eval_subs(self, old, new):
x = self.x
if old.has(x):
return self
def _eval_as_leading_term(self, x, logx=None, cdir=0):
for t in self:
if t is not S.Zero:
return t
def _eval_derivative(self, x):
f = self.function.diff(x)
ind = self.ind.diff(x)
pow_xk = self._get_pow_x(self.xk.formula)
ak = self.ak
k = ak.variables[0]
if ak.formula.has(x):
form = []
for e, c in ak.formula.args:
temp = S.Zero
for t in Add.make_args(e):
pow_x = self._get_pow_x(t)
temp += t * (pow_xk + pow_x)
form.append((temp, c))
form = Piecewise(*form)
ak = sequence(form.subs(k, k + 1), (k, ak.start - 1, ak.stop))
else:
ak = sequence((ak.formula * pow_xk).subs(k, k + 1),
(k, ak.start - 1, ak.stop))
return self.func(f, self.x, self.x0, self.dir, (ak, self.xk, ind))
def integrate(self, x=None, **kwargs):
"""
Integrate Formal Power Series.
Examples
========
>>> from sympy import fps, sin, integrate
>>> from sympy.abc import x
>>> f = fps(sin(x))
>>> f.integrate(x).truncate()
-1 + x**2/2 - x**4/24 + O(x**6)
>>> integrate(f, (x, 0, 1))
1 - cos(1)
"""
from sympy.integrals import integrate
if x is None:
x = self.x
elif iterable(x):
return integrate(self.function, x)
f = integrate(self.function, x)
ind = integrate(self.ind, x)
ind += (f - ind).limit(x, 0) # constant of integration
pow_xk = self._get_pow_x(self.xk.formula)
ak = self.ak
k = ak.variables[0]
if ak.formula.has(x):
form = []
for e, c in ak.formula.args:
temp = S.Zero
for t in Add.make_args(e):
pow_x = self._get_pow_x(t)
temp += t / (pow_xk + pow_x + 1)
form.append((temp, c))
form = Piecewise(*form)
ak = sequence(form.subs(k, k - 1), (k, ak.start + 1, ak.stop))
else:
ak = sequence((ak.formula / (pow_xk + 1)).subs(k, k - 1),
(k, ak.start + 1, ak.stop))
return self.func(f, self.x, self.x0, self.dir, (ak, self.xk, ind))
def product(self, other, x=None, n=6):
"""
Multiplies two Formal Power Series, using discrete convolution and
return the truncated terms upto specified order.
Parameters
==========
n : Number, optional
Specifies the order of the term up to which the polynomial should
be truncated.
Examples
========
>>> from sympy import fps, sin, exp
>>> from sympy.abc import x
>>> f1 = fps(sin(x))
>>> f2 = fps(exp(x))
>>> f1.product(f2, x).truncate(4)
x + x**2 + x**3/3 + O(x**4)
See Also
========
sympy.discrete.convolutions
sympy.series.formal.FormalPowerSeriesProduct
"""
if n is None:
return iter(self)
other = sympify(other)
if not isinstance(other, FormalPowerSeries):
raise ValueError("Both series should be an instance of FormalPowerSeries"
" class.")
if self.dir != other.dir:
raise ValueError("Both series should be calculated from the"
" same direction.")
elif self.x0 != other.x0:
raise ValueError("Both series should be calculated about the"
" same point.")
elif self.x != other.x:
raise ValueError("Both series should have the same symbol.")
return FormalPowerSeriesProduct(self, other)
def coeff_bell(self, n):
r"""
self.coeff_bell(n) returns a sequence of Bell polynomials of the second kind.
Note that ``n`` should be a integer.
The second kind of Bell polynomials (are sometimes called "partial" Bell
polynomials or incomplete Bell polynomials) are defined as
.. math::
B_{n,k}(x_1, x_2,\dotsc x_{n-k+1}) =
\sum_{j_1+j_2+j_2+\dotsb=k \atop j_1+2j_2+3j_2+\dotsb=n}
\frac{n!}{j_1!j_2!\dotsb j_{n-k+1}!}
\left(\frac{x_1}{1!} \right)^{j_1}
\left(\frac{x_2}{2!} \right)^{j_2} \dotsb
\left(\frac{x_{n-k+1}}{(n-k+1)!} \right) ^{j_{n-k+1}}.
* ``bell(n, k, (x1, x2, ...))`` gives Bell polynomials of the second kind,
`B_{n,k}(x_1, x_2, \dotsc, x_{n-k+1})`.
See Also
========
sympy.functions.combinatorial.numbers.bell
"""
inner_coeffs = [bell(n, j, tuple(self.bell_coeff_seq[:n-j+1])) for j in range(1, n+1)]
k = Dummy('k')
return sequence(tuple(inner_coeffs), (k, 1, oo))
def compose(self, other, x=None, n=6):
r"""
Returns the truncated terms of the formal power series of the composed function,
up to specified ``n``.
Explanation
===========
If ``f`` and ``g`` are two formal power series of two different functions,
then the coefficient sequence ``ak`` of the composed formal power series `fp`
will be as follows.
.. math::
\sum\limits_{k=0}^{n} b_k B_{n,k}(x_1, x_2, \dotsc, x_{n-k+1})
Parameters
==========
n : Number, optional
Specifies the order of the term up to which the polynomial should
be truncated.
Examples
========
>>> from sympy import fps, sin, exp
>>> from sympy.abc import x
>>> f1 = fps(exp(x))
>>> f2 = fps(sin(x))
>>> f1.compose(f2, x).truncate()
1 + x + x**2/2 - x**4/8 - x**5/15 + O(x**6)
>>> f1.compose(f2, x).truncate(8)
1 + x + x**2/2 - x**4/8 - x**5/15 - x**6/240 + x**7/90 + O(x**8)
See Also
========
sympy.functions.combinatorial.numbers.bell
sympy.series.formal.FormalPowerSeriesCompose
References
==========
.. [1] Comtet, Louis: Advanced combinatorics; the art of finite and infinite expansions. Reidel, 1974.
"""
if n is None:
return iter(self)
other = sympify(other)
if not isinstance(other, FormalPowerSeries):
raise ValueError("Both series should be an instance of FormalPowerSeries"
" class.")
if self.dir != other.dir:
raise ValueError("Both series should be calculated from the"
" same direction.")
elif self.x0 != other.x0:
raise ValueError("Both series should be calculated about the"
" same point.")
elif self.x != other.x:
raise ValueError("Both series should have the same symbol.")
if other._eval_term(0).as_coeff_mul(other.x)[0] is not S.Zero:
raise ValueError("The formal power series of the inner function should not have any "
"constant coefficient term.")
return FormalPowerSeriesCompose(self, other)
def inverse(self, x=None, n=6):
r"""
Returns the truncated terms of the inverse of the formal power series,
up to specified ``n``.
Explanation
===========
If ``f`` and ``g`` are two formal power series of two different functions,
then the coefficient sequence ``ak`` of the composed formal power series ``fp``
will be as follows.
.. math::
\sum\limits_{k=0}^{n} (-1)^{k} x_0^{-k-1} B_{n,k}(x_1, x_2, \dotsc, x_{n-k+1})
Parameters
==========
n : Number, optional
Specifies the order of the term up to which the polynomial should
be truncated.
Examples
========
>>> from sympy import fps, exp, cos
>>> from sympy.abc import x
>>> f1 = fps(exp(x))
>>> f2 = fps(cos(x))
>>> f1.inverse(x).truncate()
1 - x + x**2/2 - x**3/6 + x**4/24 - x**5/120 + O(x**6)
>>> f2.inverse(x).truncate(8)
1 + x**2/2 + 5*x**4/24 + 61*x**6/720 + O(x**8)
See Also
========
sympy.functions.combinatorial.numbers.bell
sympy.series.formal.FormalPowerSeriesInverse
References
==========
.. [1] Comtet, Louis: Advanced combinatorics; the art of finite and infinite expansions. Reidel, 1974.
"""
if n is None:
return iter(self)
if self._eval_term(0).is_zero:
raise ValueError("Constant coefficient should exist for an inverse of a formal"
" power series to exist.")
return FormalPowerSeriesInverse(self)
def __add__(self, other):
other = sympify(other)
if isinstance(other, FormalPowerSeries):
if self.dir != other.dir:
raise ValueError("Both series should be calculated from the"
" same direction.")
elif self.x0 != other.x0:
raise ValueError("Both series should be calculated about the"
" same point.")
x, y = self.x, other.x
f = self.function + other.function.subs(y, x)
if self.x not in f.free_symbols:
return f
ak = self.ak + other.ak
if self.ak.start > other.ak.start:
seq = other.ak
s, e = other.ak.start, self.ak.start
else:
seq = self.ak
s, e = self.ak.start, other.ak.start
save = Add(*[z[0]*z[1] for z in zip(seq[0:(e - s)], self.xk[s:e])])
ind = self.ind + other.ind + save
return self.func(f, x, self.x0, self.dir, (ak, self.xk, ind))
elif not other.has(self.x):
f = self.function + other
ind = self.ind + other
return self.func(f, self.x, self.x0, self.dir,
(self.ak, self.xk, ind))
return Add(self, other)
def __radd__(self, other):
return self.__add__(other)
def __neg__(self):
return self.func(-self.function, self.x, self.x0, self.dir,
(-self.ak, self.xk, -self.ind))
def __sub__(self, other):
return self.__add__(-other)
def __rsub__(self, other):
return (-self).__add__(other)
def __mul__(self, other):
other = sympify(other)
if other.has(self.x):
return Mul(self, other)
f = self.function * other
ak = self.ak.coeff_mul(other)
ind = self.ind * other
return self.func(f, self.x, self.x0, self.dir, (ak, self.xk, ind))
def __rmul__(self, other):
return self.__mul__(other)
class FiniteFormalPowerSeries(FormalPowerSeries):
"""Base Class for Product, Compose and Inverse classes"""
def __init__(self, *args):
pass
@property
def ffps(self):
return self.args[0]
@property
def gfps(self):
return self.args[1]
@property
def f(self):
return self.ffps.function
@property
def g(self):
return self.gfps.function
@property
def infinite(self):
raise NotImplementedError("No infinite version for an object of"
" FiniteFormalPowerSeries class.")
def _eval_terms(self, n):
raise NotImplementedError("(%s)._eval_terms()" % self)
def _eval_term(self, pt):
raise NotImplementedError("By the current logic, one can get terms"
"upto a certain order, instead of getting term by term.")
def polynomial(self, n):
return self._eval_terms(n)
def truncate(self, n=6):
ffps = self.ffps
pt_xk = ffps.xk.coeff(n)
x, x0 = ffps.x, ffps.x0
return self.polynomial(n) + Order(pt_xk, (x, x0))
def _eval_derivative(self, x):
raise NotImplementedError
def integrate(self, x):
raise NotImplementedError
class FormalPowerSeriesProduct(FiniteFormalPowerSeries):
"""Represents the product of two formal power series of two functions.
Explanation
===========
No computation is performed. Terms are calculated using a term by term logic,
instead of a point by point logic.
There are two differences between a :obj:`FormalPowerSeries` object and a
:obj:`FormalPowerSeriesProduct` object. The first argument contains the two
functions involved in the product. Also, the coefficient sequence contains
both the coefficient sequence of the formal power series of the involved functions.
See Also
========
sympy.series.formal.FormalPowerSeries
sympy.series.formal.FiniteFormalPowerSeries
"""
def __init__(self, *args):
ffps, gfps = self.ffps, self.gfps
k = ffps.ak.variables[0]
self.coeff1 = sequence(ffps.ak.formula, (k, 0, oo))
k = gfps.ak.variables[0]
self.coeff2 = sequence(gfps.ak.formula, (k, 0, oo))
@property
def function(self):
"""Function of the product of two formal power series."""
return self.f * self.g
def _eval_terms(self, n):
"""
Returns the first ``n`` terms of the product formal power series.
Term by term logic is implemented here.
Examples
========
>>> from sympy import fps, sin, exp
>>> from sympy.abc import x
>>> f1 = fps(sin(x))
>>> f2 = fps(exp(x))
>>> fprod = f1.product(f2, x)
>>> fprod._eval_terms(4)
x**3/3 + x**2 + x
See Also
========
sympy.series.formal.FormalPowerSeries.product
"""
coeff1, coeff2 = self.coeff1, self.coeff2
aks = convolution(coeff1[:n], coeff2[:n])
terms = []
for i in range(0, n):
terms.append(aks[i] * self.ffps.xk.coeff(i))
return Add(*terms)
class FormalPowerSeriesCompose(FiniteFormalPowerSeries):
"""
Represents the composed formal power series of two functions.
Explanation
===========
No computation is performed. Terms are calculated using a term by term logic,
instead of a point by point logic.
There are two differences between a :obj:`FormalPowerSeries` object and a
:obj:`FormalPowerSeriesCompose` object. The first argument contains the outer
function and the inner function involved in the omposition. Also, the
coefficient sequence contains the generic sequence which is to be multiplied
by a custom ``bell_seq`` finite sequence. The finite terms will then be added up to
get the final terms.
See Also
========
sympy.series.formal.FormalPowerSeries
sympy.series.formal.FiniteFormalPowerSeries
"""
@property
def function(self):
"""Function for the composed formal power series."""
f, g, x = self.f, self.g, self.ffps.x
return f.subs(x, g)
def _eval_terms(self, n):
"""
Returns the first `n` terms of the composed formal power series.
Term by term logic is implemented here.
Explanation
===========
The coefficient sequence of the :obj:`FormalPowerSeriesCompose` object is the generic sequence.
It is multiplied by ``bell_seq`` to get a sequence, whose terms are added up to get
the final terms for the polynomial.
Examples
========
>>> from sympy import fps, sin, exp
>>> from sympy.abc import x
>>> f1 = fps(exp(x))
>>> f2 = fps(sin(x))
>>> fcomp = f1.compose(f2, x)
>>> fcomp._eval_terms(6)
-x**5/15 - x**4/8 + x**2/2 + x + 1
>>> fcomp._eval_terms(8)
x**7/90 - x**6/240 - x**5/15 - x**4/8 + x**2/2 + x + 1
See Also
========
sympy.series.formal.FormalPowerSeries.compose
sympy.series.formal.FormalPowerSeries.coeff_bell
"""
ffps, gfps = self.ffps, self.gfps
terms = [ffps.zero_coeff()]
for i in range(1, n):
bell_seq = gfps.coeff_bell(i)
seq = (ffps.bell_coeff_seq * bell_seq)
terms.append(Add(*(seq[:i])) / ffps.fact_seq[i-1] * ffps.xk.coeff(i))
return Add(*terms)
class FormalPowerSeriesInverse(FiniteFormalPowerSeries):
"""
Represents the Inverse of a formal power series.
Explanation
===========
No computation is performed. Terms are calculated using a term by term logic,
instead of a point by point logic.
There is a single difference between a :obj:`FormalPowerSeries` object and a
:obj:`FormalPowerSeriesInverse` object. The coefficient sequence contains the
generic sequence which is to be multiplied by a custom ``bell_seq`` finite sequence.
The finite terms will then be added up to get the final terms.
See Also
========
sympy.series.formal.FormalPowerSeries
sympy.series.formal.FiniteFormalPowerSeries
"""
def __init__(self, *args):
ffps = self.ffps
k = ffps.xk.variables[0]
inv = ffps.zero_coeff()
inv_seq = sequence(inv ** (-(k + 1)), (k, 1, oo))
self.aux_seq = ffps.sign_seq * ffps.fact_seq * inv_seq
@property
def function(self):
"""Function for the inverse of a formal power series."""
f = self.f
return 1 / f
@property
def g(self):
raise ValueError("Only one function is considered while performing"
"inverse of a formal power series.")
@property
def gfps(self):
raise ValueError("Only one function is considered while performing"
"inverse of a formal power series.")
def _eval_terms(self, n):
"""
Returns the first ``n`` terms of the composed formal power series.
Term by term logic is implemented here.
Explanation
===========
The coefficient sequence of the `FormalPowerSeriesInverse` object is the generic sequence.
It is multiplied by ``bell_seq`` to get a sequence, whose terms are added up to get
the final terms for the polynomial.
Examples
========
>>> from sympy import fps, exp, cos
>>> from sympy.abc import x
>>> f1 = fps(exp(x))
>>> f2 = fps(cos(x))
>>> finv1, finv2 = f1.inverse(), f2.inverse()
>>> finv1._eval_terms(6)
-x**5/120 + x**4/24 - x**3/6 + x**2/2 - x + 1
>>> finv2._eval_terms(8)
61*x**6/720 + 5*x**4/24 + x**2/2 + 1
See Also
========
sympy.series.formal.FormalPowerSeries.inverse
sympy.series.formal.FormalPowerSeries.coeff_bell
"""
ffps = self.ffps
terms = [ffps.zero_coeff()]
for i in range(1, n):
bell_seq = ffps.coeff_bell(i)
seq = (self.aux_seq * bell_seq)
terms.append(Add(*(seq[:i])) / ffps.fact_seq[i-1] * ffps.xk.coeff(i))
return Add(*terms)
def fps(f, x=None, x0=0, dir=1, hyper=True, order=4, rational=True, full=False):
"""
Generates Formal Power Series of ``f``.
Explanation
===========
Returns the formal series expansion of ``f`` around ``x = x0``
with respect to ``x`` in the form of a ``FormalPowerSeries`` object.
Formal Power Series is represented using an explicit formula
computed using different algorithms.
See :func:`compute_fps` for the more details regarding the computation
of formula.
Parameters
==========
x : Symbol, optional
If x is None and ``f`` is univariate, the univariate symbols will be
supplied, otherwise an error will be raised.
x0 : number, optional
Point to perform series expansion about. Default is 0.
dir : {1, -1, '+', '-'}, optional
If dir is 1 or '+' the series is calculated from the right and
for -1 or '-' the series is calculated from the left. For smooth
functions this flag will not alter the results. Default is 1.
hyper : {True, False}, optional
Set hyper to False to skip the hypergeometric algorithm.
By default it is set to False.
order : int, optional
Order of the derivative of ``f``, Default is 4.
rational : {True, False}, optional
Set rational to False to skip rational algorithm. By default it is set
to True.
full : {True, False}, optional
Set full to True to increase the range of rational algorithm.
See :func:`rational_algorithm` for details. By default it is set to
False.
Examples
========
>>> from sympy import fps, ln, atan, sin
>>> from sympy.abc import x, n
Rational Functions
>>> fps(ln(1 + x)).truncate()
x - x**2/2 + x**3/3 - x**4/4 + x**5/5 + O(x**6)
>>> fps(atan(x), full=True).truncate()
x - x**3/3 + x**5/5 + O(x**6)
Symbolic Functions
>>> fps(x**n*sin(x**2), x).truncate(8)
-x**(n + 6)/6 + x**(n + 2) + O(x**(n + 8))
See Also
========
sympy.series.formal.FormalPowerSeries
sympy.series.formal.compute_fps
"""
f = sympify(f)
if x is None:
free = f.free_symbols
if len(free) == 1:
x = free.pop()
elif not free:
return f
else:
raise NotImplementedError("multivariate formal power series")
result = compute_fps(f, x, x0, dir, hyper, order, rational, full)
if result is None:
return f
return FormalPowerSeries(f, x, x0, dir, result)