Files
2026-07-13 12:32:53 +08:00

284 lines
7.6 KiB
Python

import cmath
import functools
import inspect
import math
import sys
from . import function_docs, libfp, libmp
from .ctx_base import StandardBaseContext
from .libmp import int_types, mpf_bernoulli, to_float
class FPContext(StandardBaseContext):
"""
Context for fast low-precision arithmetic (usually, 53-bit precision,
giving at most about 15 decimal digits), using Python's builtin float and
complex types.
"""
def __init__(ctx):
super().__init__()
ctx.pretty = False
ctx._init_aliases()
NoConvergence = libmp.NoConvergence
@property
def prec(ctx):
return sys.float_info.mant_dig
@prec.setter
def prec(ctx, p):
return
@property
def dps(ctx):
return sys.float_info.dig
@dps.setter
def dps(ctx, p):
return
_fixed_precision = True
zero = 0.0
one = 1.0
eps = sys.float_info.epsilon
inf = libfp.INF
ninf = -math.inf
nan = math.nan
j = 1j
# Called by SpecialFunctions.__init__()
@classmethod
def _wrap_specfun(cls, name, f, wrap):
if wrap:
def f_wrapped(ctx, *args, **kwargs):
convert = ctx.convert
args = [convert(a) for a in args]
return f(ctx, *args, **kwargs)
else:
f_wrapped = f
f_wrapped.__doc__ = function_docs.__dict__.get(name, f.__doc__)
try:
f_wrapped.__signature__ = inspect.signature(f)
except ValueError: # pragma: no cover
pass
f_wrapped.__name__ = f.__name__
setattr(cls, name, f_wrapped)
@functools.lru_cache
def bernoulli(ctx, n, plus=False):
return to_float(mpf_bernoulli(n, ctx.prec, 'n', plus=plus), strict=True)
pi = libfp.pi
e = math.e
euler = libfp.euler
sqrt2 = 1.4142135623730950488
sqrt5 = 2.2360679774997896964
phi = 1.6180339887498948482
ln2 = 0.69314718055994530942
ln10 = 2.302585092994045684
euler = libfp.euler
catalan = 0.91596559417721901505
khinchin = 2.6854520010653064453
apery = 1.2020569031595942854
glaisher = 1.2824271291006226369
absmin = absmax = abs
def isnan(ctx, x):
return x != x
def isinf(ctx, x):
return abs(x) == libfp.INF
def isfinite(ctx, x):
if type(x) is complex:
return all(map(math.isfinite, [x.real, x.imag]))
return math.isfinite(x)
def isnormal(ctx, x):
if type(x) is complex:
return ctx.isnormal(abs(x))
# XXX: can use math.isnormal() on Python 3.15+
return bool(x) and math.isfinite(x) and abs(x) >= sys.float_info.min
def isnpint(ctx, x):
if type(x) is complex:
if x.imag:
return False
x = x.real
return math.isfinite(x) and x <= 0.0 and round(x) == x
mpf = float
mpc = complex
def convert(ctx, x):
try:
return float(x)
except:
return complex(x)
power = staticmethod(libfp.pow)
sqrt = staticmethod(libfp.sqrt)
exp = staticmethod(libfp.exp)
ln = log = staticmethod(libfp.log)
cos = staticmethod(libfp.cos)
sin = staticmethod(libfp.sin)
tan = staticmethod(libfp.tan)
cos_sin = staticmethod(libfp.cos_sin)
acos = staticmethod(libfp.acos)
asin = staticmethod(libfp.asin)
atan = staticmethod(libfp.atan)
cosh = staticmethod(libfp.cosh)
sinh = staticmethod(libfp.sinh)
tanh = staticmethod(libfp.tanh)
acosh = staticmethod(libfp.acosh)
asinh = staticmethod(libfp.asinh)
atanh = staticmethod(libfp.atanh)
gamma = staticmethod(libfp.gamma)
rgamma = staticmethod(libfp.rgamma)
fac = factorial = staticmethod(libfp.factorial)
floor = staticmethod(libfp.floor)
ceil = staticmethod(libfp.ceil)
cospi = staticmethod(libfp.cospi)
sinpi = staticmethod(libfp.sinpi)
cbrt = staticmethod(libfp.cbrt)
_nthroot = staticmethod(libfp.nthroot)
_ei = staticmethod(libfp.ei)
_e1 = staticmethod(libfp.e1)
_zeta = _zeta_int = staticmethod(libfp.zeta)
arg = staticmethod(cmath.phase)
loggamma = staticmethod(libfp.loggamma)
def expj(ctx, x):
return ctx.exp(ctx.j*x)
def expjpi(ctx, x):
return ctx.exp(ctx.j*ctx.pi*x)
ldexp = math.ldexp
frexp = math.frexp
hypot = math.hypot
def mag(ctx, z):
if z:
n, e = ctx.frexp(abs(z))
if e:
return e
return ctx.convert(n)
return ctx.ninf
def isint(ctx, z):
if z.imag:
return False
z = z.real
try:
return z == int(z)
except:
return False
def nint_distance(ctx, z):
n = round(z.real)
if n == z:
return n, ctx.ninf
return n, ctx.mag(abs(z-n))
def _convert_param(ctx, z):
if type(z) is tuple:
p, q = z
return ctx.mpf(p / q), 'R'
intz = int(z.real)
if z == intz:
return intz, 'Z'
if not z.imag:
return ctx.mpf(z), 'R'
return ctx.mpc(z), 'C'
def _is_real_type(ctx, z):
return isinstance(z, float) or isinstance(z, int_types)
def _is_complex_type(ctx, z):
return isinstance(z, complex)
def hypsum(ctx, p, q, flags, coeffs, z, maxterms=6000, **kwargs):
for i, c in enumerate(coeffs[p:], start=p):
if flags[i] == 'Z':
if c <= 0:
ok = False
for ii, cc in enumerate(coeffs[:p]):
# Note: c <= cc or c < cc, depending on convention
if flags[ii] == 'Z' and cc <= 0 and c <= cc:
ok = True
if not ok:
raise ZeroDivisionError("pole in hypergeometric series")
num = range(p)
den = range(p,p+q)
if ctx.isinf(z):
n = max(((n, c) for n, c in enumerate(coeffs[:p])
if flags[n] == 'Z' and c < 0), default=(-1, 0),
key=lambda x: x[1])[0]
if n >= 0:
n = -coeffs[n]
t = z**n
for k in range(n):
for i in num: t *= (coeffs[i]+k)
for i in den: t /= (coeffs[i]+k)
t /= (k+1)
return t
tol = ctx.eps
s = t = 1.0
k = 0
while 1:
for i in num: t *= (coeffs[i]+k)
try:
for i in den: t /= (coeffs[i]+k)
except ZeroDivisionError:
assert not t # poles are handled above
return s
k += 1; t /= k; t *= z; s += t
if abs(t) < tol:
return s
if k > maxterms:
raise ctx.NoConvergence
atan2 = staticmethod(math.atan2)
def psi(ctx, m, z):
m = int(m)
if m == 0:
return ctx.digamma(z)
return (-1)**(m+1) * ctx.fac(m) * ctx.zeta(m+1, z)
digamma = staticmethod(libfp.digamma)
def harmonic(ctx, x):
x = ctx.convert(x)
if x == 0 or x == 1:
return x
return ctx.digamma(x+1) + ctx.euler
nstr = str
def to_fixed(ctx, x, prec):
return int(math.ldexp(x, prec))
def rand(ctx):
import random
return random.random()
_erf = staticmethod(math.erf)
_erfc = staticmethod(math.erfc)
def sum_accurately(ctx, terms, check_step=1):
s = ctx.zero
k = 0
for term in terms():
s += term
if (not k % check_step) and term:
if abs(term) <= 1e-18*abs(s):
break
k += 1
return s