luke
/
OpenRAM
mirror of https://github.com/VLSIDA/OpenRAM.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
OpenRAM/compiler/gdsMill/mpmath/libmp/libelefun.py

1596 lines
50 KiB
Python
Raw Normal View History

RELEASE 1.0
2016-11-08 18:57:35 +01:00
"""
This module implements computation of elementary transcendental
functions (powers, logarithms, trigonometric and hyperbolic
functions, inverse trigonometric and hyperbolic) for real
floating-point numbers.
For complex and interval implementations of the same functions,
see libmpc and libmpi.
"""
import math
from bisect import bisect
from backend import MPZ, MPZ_ZERO, MPZ_ONE, MPZ_TWO, MPZ_FIVE
from libmpf import (
round_floor, round_ceiling, round_down, round_up,
round_nearest, round_fast,
ComplexResult,
bitcount, bctable, lshift, rshift, giant_steps, sqrt_fixed,
from_int, to_int, from_man_exp, to_fixed, to_float, from_float,
normalize,
fzero, fone, fnone, fhalf, finf, fninf, fnan,
mpf_cmp, mpf_sign, mpf_abs,
mpf_pos, mpf_neg, mpf_add, mpf_sub, mpf_mul, mpf_div, mpf_shift,
mpf_rdiv_int, mpf_pow_int, mpf_sqrt,
reciprocal_rnd, negative_rnd, mpf_perturb,
isqrt_fast
)
from libintmath import ifib
#----------------------------------------------------------------------------#
# #
# Elementary mathematical constants #
# #
#----------------------------------------------------------------------------#
def constant_memo(f):
"""
Decorator for caching computed values of mathematical
constants. This decorator should be applied to a
function taking a single argument prec as input and
returning a fixed-point value with the given precision.
"""
f.memo_prec = -1
f.memo_val = None
def g(prec, **kwargs):
memo_prec = f.memo_prec
if prec <= memo_prec:
return f.memo_val >> (memo_prec-prec)
newprec = int(prec*1.05+10)
f.memo_val = f(newprec, **kwargs)
f.memo_prec = newprec
return f.memo_val >> (newprec-prec)
g.__name__ = f.__name__
g.__doc__ = f.__doc__
return g
def def_mpf_constant(fixed):
"""
Create a function that computes the mpf value for a mathematical
constant, given a function that computes the fixed-point value.
Assumptions: the constant is positive and has magnitude ~= 1;
the fixed-point function rounds to floor.
"""
def f(prec, rnd=round_fast):
wp = prec + 20
v = fixed(wp)
if rnd in (round_up, round_ceiling):
v += 1
return normalize(0, v, -wp, bitcount(v), prec, rnd)
f.__doc__ = fixed.__doc__
return f
def bsp_acot(q, a, b, hyperbolic):
if b - a == 1:
a1 = MPZ(2*a + 3)
if hyperbolic or a&1:
return MPZ_ONE, a1 * q**2, a1
else:
return -MPZ_ONE, a1 * q**2, a1
m = (a+b)//2
p1, q1, r1 = bsp_acot(q, a, m, hyperbolic)
p2, q2, r2 = bsp_acot(q, m, b, hyperbolic)
return q2*p1 + r1*p2, q1*q2, r1*r2
# the acoth(x) series converges like the geometric series for x^2
# N = ceil(p*log(2)/(2*log(x)))
def acot_fixed(a, prec, hyperbolic):
"""
Compute acot(a) or acoth(a) for an integer a with binary splitting; see
http://numbers.computation.free.fr/Constants/Algorithms/splitting.html
"""
N = int(0.35 * prec/math.log(a) + 20)
p, q, r = bsp_acot(a, 0,N, hyperbolic)
return ((p+q)<<prec)//(q*a)
def machin(coefs, prec, hyperbolic=False):
"""
Evaluate a Machin-like formula, i.e., a linear combination of
acot(n) or acoth(n) for specific integer values of n, using fixed-
point arithmetic. The input should be a list [(c, n), ...], giving
c*acot[h](n) + ...
"""
extraprec = 10
s = MPZ_ZERO
for a, b in coefs:
s += MPZ(a) * acot_fixed(MPZ(b), prec+extraprec, hyperbolic)
return (s >> extraprec)
# Logarithms of integers are needed for various computations involving
# logarithms, powers, radix conversion, etc
@constant_memo
def ln2_fixed(prec):
"""
Computes ln(2). This is done with a hyperbolic Machin-type formula,
with binary splitting at high precision.
"""
return machin([(18, 26), (-2, 4801), (8, 8749)], prec, True)
@constant_memo
def ln10_fixed(prec):
"""
Computes ln(10). This is done with a hyperbolic Machin-type formula.
"""
return machin([(46, 31), (34, 49), (20, 161)], prec, True)
"""
For computation of pi, we use the Chudnovsky series:
oo
___ k
1 \ (-1) (6 k)! (A + B k)
----- = ) -----------------------
12 pi /___ 3 3k+3/2
(3 k)! (k!) C
k = 0
where A, B, and C are certain integer constants. This series adds roughly
14 digits per term. Note that C^(3/2) can be extracted so that the
series contains only rational terms. This makes binary splitting very
efficient.
The recurrence formulas for the binary splitting were taken from
ftp://ftp.gmplib.org/pub/src/gmp-chudnovsky.c
Previously, Machin's formula was used at low precision and the AGM iteration
was used at high precision. However, the Chudnovsky series is essentially as
fast as the Machin formula at low precision and in practice about 3x faster
than the AGM at high precision (despite theoretically having a worse
asymptotic complexity), so there is no reason not to use it in all cases.
"""
# Constants in Chudnovsky's series
CHUD_A = MPZ(13591409)
CHUD_B = MPZ(545140134)
CHUD_C = MPZ(640320)
CHUD_D = MPZ(12)
def bs_chudnovsky(a, b, level, verbose):
"""
Computes the sum from a to b of the series in the Chudnovsky
formula. Returns g, p, q where p/q is the sum as an exact
fraction and g is a temporary value used to save work
for recursive calls.
"""
if b-a == 1:
g = MPZ((6*b-5)*(2*b-1)*(6*b-1))
p = b**3 * CHUD_C**3 // 24
q = (-1)**b * g * (CHUD_A+CHUD_B*b)
else:
if verbose and level < 4:
print " binary splitting", a, b
mid = (a+b)//2
g1, p1, q1 = bs_chudnovsky(a, mid, level+1, verbose)
g2, p2, q2 = bs_chudnovsky(mid, b, level+1, verbose)
p = p1*p2
g = g1*g2
q = q1*p2 + q2*g1
return g, p, q
@constant_memo
def pi_fixed(prec, verbose=False, verbose_base=None):
"""
Compute floor(pi * 2**prec) as a big integer.
This is done using Chudnovsky's series (see comments in
libelefun.py for details).
"""
# The Chudnovsky series gives 14.18 digits per term
N = int(prec/3.3219280948/14.181647462 + 2)
if verbose:
print "binary splitting with N =", N
g, p, q = bs_chudnovsky(0, N, 0, verbose)
sqrtC = isqrt_fast(CHUD_C<<(2*prec))
v = p*CHUD_C*sqrtC//((q+CHUD_A*p)*CHUD_D)
return v
def degree_fixed(prec):
return pi_fixed(prec)//180
def bspe(a, b):
"""
Sum series for exp(1)-1 between a, b, returning the result
as an exact fraction (p, q).
"""
if b-a == 1:
return MPZ_ONE, MPZ(b)
m = (a+b)//2
p1, q1 = bspe(a, m)
p2, q2 = bspe(m, b)
return p1*q2+p2, q1*q2
@constant_memo
def e_fixed(prec):
"""
Computes exp(1). This is done using the ordinary Taylor series for
exp, with binary splitting. For a description of the algorithm,
see:
http://numbers.computation.free.fr/Constants/
Algorithms/splitting.html
"""
# Slight overestimate of N needed for 1/N! < 2**(-prec)
# This could be tightened for large N.
N = int(1.1*prec/math.log(prec) + 20)
p, q = bspe(0,N)
return ((p+q)<<prec)//q
@constant_memo
def phi_fixed(prec):
"""
Computes the golden ratio, (1+sqrt(5))/2
"""
prec += 10
a = isqrt_fast(MPZ_FIVE<<(2*prec)) + (MPZ_ONE << prec)
return a >> 11
mpf_phi = def_mpf_constant(phi_fixed)
mpf_pi = def_mpf_constant(pi_fixed)
mpf_e = def_mpf_constant(e_fixed)
mpf_degree = def_mpf_constant(degree_fixed)
mpf_ln2 = def_mpf_constant(ln2_fixed)
mpf_ln10 = def_mpf_constant(ln10_fixed)
#----------------------------------------------------------------------------#
# #
# Powers #
# #
#----------------------------------------------------------------------------#
def mpf_pow(s, t, prec, rnd=round_fast):
"""
Compute s**t. Raises ComplexResult if s is negative and t is
fractional.
"""
ssign, sman, sexp, sbc = s
tsign, tman, texp, tbc = t
if ssign and texp < 0:
raise ComplexResult("negative number raised to a fractional power")
if texp >= 0:
return mpf_pow_int(s, (-1)**tsign * (tman<<texp), prec, rnd)
# s**(n/2) = sqrt(s)**n
if texp == -1:
if tman == 1:
if tsign:
return mpf_div(fone, mpf_sqrt(s, prec+10,
reciprocal_rnd[rnd]), prec, rnd)
return mpf_sqrt(s, prec, rnd)
else:
if tsign:
return mpf_pow_int(mpf_sqrt(s, prec+10,
reciprocal_rnd[rnd]), -tman, prec, rnd)
return mpf_pow_int(mpf_sqrt(s, prec+10, rnd), tman, prec, rnd)
# General formula: s**t = exp(t*log(s))
# TODO: handle rnd direction of the logarithm carefully
c = mpf_log(s, prec+10, rnd)
return mpf_exp(mpf_mul(t, c), prec, rnd)
def int_pow_fixed(y, n, prec):
"""n-th power of a fixed point number with precision prec
Returns the power in the form man, exp,
man * 2**exp ~= y**n
"""
if n == 2:
return (y*y), 0
bc = bitcount(y)
exp = 0
workprec = 2 * (prec + 4*bitcount(n) + 4)
_, pm, pe, pbc = fone
while 1:
if n & 1:
pm = pm*y
pe = pe+exp
pbc += bc - 2
pbc = pbc + bctable[int(pm >> pbc)]
if pbc > workprec:
pm = pm >> (pbc-workprec)
pe += pbc - workprec
pbc = workprec
n -= 1
if not n:
break
y = y*y
exp = exp+exp
bc = bc + bc - 2
bc = bc + bctable[int(y >> bc)]
if bc > workprec:
y = y >> (bc-workprec)
exp += bc - workprec
bc = workprec
n = n // 2
return pm, pe
# froot(s, n, prec, rnd) computes the real n-th root of a
# positive mpf tuple s.
# To compute the root we start from a 50-bit estimate for r
# generated with ordinary floating-point arithmetic, and then refine
# the value to full accuracy using the iteration
# 1 / y \
# r = --- | (n-1) * r + ---------- |
# n+1 n \ n r_n**(n-1) /
# which is simply Newton's method applied to the equation r**n = y.
# With giant_steps(start, prec+extra) = [p0,...,pm, prec+extra]
# and y = man * 2**-shift one has
# (man * 2**exp)**(1/n) =
# y**(1/n) * 2**(start-prec/n) * 2**(p0-start) * ... * 2**(prec+extra-pm) *
# 2**((exp+shift-(n-1)*prec)/n -extra))
# The last factor is accounted for in the last line of froot.
def nthroot_fixed(y, n, prec, exp1):
start = 50
try:
y1 = rshift(y, prec - n*start)
r = MPZ(int(y1**(1.0/n)))
except OverflowError:
y1 = from_int(y1, start)
fn = from_int(n)
fn = mpf_rdiv_int(1, fn, start)
r = mpf_pow(y1, fn, start)
r = to_int(r)
extra = 10
extra1 = n
prevp = start
for p in giant_steps(start, prec+extra):
pm, pe = int_pow_fixed(r, n-1, prevp)
r2 = rshift(pm, (n-1)*prevp - p - pe - extra1)
B = lshift(y, 2*p-prec+extra1)//r2
r = (B + (n-1) * lshift(r, p-prevp))//n
prevp = p
return r
def mpf_nthroot(s, n, prec, rnd=round_fast):
"""nth-root of a positive number
Use the Newton method when faster, otherwise use x**(1/n)
"""
sign, man, exp, bc = s
if sign:
raise ComplexResult("nth root of a negative number")
if not man:
if s == fnan:
return fnan
if s == fzero:
if n > 0:
return fzero
if n == 0:
return fone
return finf
# Infinity
if not n:
return fnan
if n < 0:
return fzero
return finf
flag_inverse = False
if n < 2:
if n == 0:
return fone
if n == 1:
return mpf_pos(s, prec, rnd)
if n == -1:
return mpf_div(fone, s, prec, rnd)
# n < 0
rnd = reciprocal_rnd[rnd]
flag_inverse = True
extra_inverse = 5
prec += extra_inverse
n = -n
if n > 20 and (n >= 20000 or prec < int(233 + 28.3 * n**0.62)):
prec2 = prec + 10
fn = from_int(n)
nth = mpf_rdiv_int(1, fn, prec2)
r = mpf_pow(s, nth, prec2, rnd)
s = normalize(r[0], r[1], r[2], r[3], prec, rnd)
if flag_inverse:
return mpf_div(fone, s, prec-extra_inverse, rnd)
else:
return s
# Convert to a fixed-point number with prec2 bits.
prec2 = prec + 2*n - (prec%n)
# a few tests indicate that
# for 10 < n < 10**4 a bit more precision is needed
if n > 10:
prec2 += prec2//10
prec2 = prec2 - prec2%n
# Mantissa may have more bits than we need. Trim it down.
shift = bc - prec2
# Adjust exponents to make prec2 and exp+shift multiples of n.
sign1 = 0
es = exp+shift
if es < 0:
sign1 = 1
es = -es
if sign1:
shift += es%n
else:
shift -= es%n
man = rshift(man, shift)
extra = 10
exp1 = ((exp+shift-(n-1)*prec2)//n) - extra
rnd_shift = 0
if flag_inverse:
if rnd == 'u' or rnd == 'c':
rnd_shift = 1
else:
if rnd == 'd' or rnd == 'f':
rnd_shift = 1
man = nthroot_fixed(man+rnd_shift, n, prec2, exp1)
s = from_man_exp(man, exp1, prec, rnd)
if flag_inverse:
return mpf_div(fone, s, prec-extra_inverse, rnd)
else:
return s
def mpf_cbrt(s, prec, rnd=round_fast):
"""cubic root of a positive number"""
return mpf_nthroot(s, 3, prec, rnd)
#----------------------------------------------------------------------------#
# #
# Logarithms #
# #
#----------------------------------------------------------------------------#
# Fast sequential integer logarithms are required for various series
# computations related to zeta functions, so we cache them
# TODO: can this be done better?
MAX_LOG_INT_CACHE = 2000
log_int_cache = {}
def log_int_fixed(n, prec):
if n in log_int_cache:
value, vprec = log_int_cache[n]
if vprec >= prec:
return value >> (vprec - prec)
extra = 30
vprec = prec + extra
v = to_fixed(mpf_log(from_int(n), vprec+5), vprec)
if n < MAX_LOG_INT_CACHE:
log_int_cache[n] = (v, vprec)
return v >> extra
# Use Taylor series with caching up to this prec
LOG_TAYLOR_PREC = 2500
# Cache log values in steps of size 2^-N
LOG_TAYLOR_SHIFT = 9
# prec/size ratio of x for fastest convergence in AGM formula
LOG_AGM_MAG_PREC_RATIO = 20
log_taylor_cache = {}
# ~= next power of two + 20
cache_prec_steps = [22,22]
for k in xrange(1, bitcount(LOG_TAYLOR_PREC)+1):
cache_prec_steps += [min(2**k,LOG_TAYLOR_PREC)+20] * 2**(k-1)
def agm_fixed(a, b, prec):
"""
Fixed-point computation of agm(a,b), assuming
a, b both close to unit magnitude.
"""
i = 0
while 1:
anew = (a+b)>>1
if i > 4 and abs(a-anew) < 8:
return a
b = isqrt_fast(a*b)
a = anew
i += 1
return a
def log_agm(x, prec):
"""
Fixed-point computation of -log(x) = log(1/x), suitable
for large precision. It is required that 0 < x < 1. The
algorithm used is the Sasaki-Kanada formula
-log(x) = pi/agm(theta2(x)^2,theta3(x)^2). [1]
For faster convergence in the theta functions, x should
be chosen closer to 0.
Guard bits must be added by the caller.
HYPOTHESIS: if x = 2^(-n), n bits need to be added to
account for the truncation to a fixed-point number,
and this is the only significant cancellation error.
The number of bits lost to roundoff is small and can be
considered constant.
[1] Richard P. Brent, "Fast Algorithms for High-Precision
Computation of Elementary Functions (extended abstract)",
http://wwwmaths.anu.edu.au/~brent/pd/RNC7-Brent.pdf
"""
x2 = (x*x) >> prec
# Compute jtheta2(x)**2
s = a = b = x2
while a:
b = (b*x2) >> prec
a = (a*b) >> prec
s += a
s += (MPZ_ONE<<prec)
s = (s*s)>>(prec-2)
s = (s*isqrt_fast(x<<prec))>>prec
# Compute jtheta3(x)**2
t = a = b = x
while a:
b = (b*x2) >> prec
a = (a*b) >> prec
t += a
t = (MPZ_ONE<<prec) + (t<<1)
t = (t*t)>>prec
# Final formula
p = agm_fixed(s, t, prec)
return (pi_fixed(prec) << prec) // p
def log_taylor(x, prec, r=0):
"""
Fixed-point calculation of log(x). It is assumed that x is close
enough to 1 for the Taylor series to converge quickly. Convergence
can be improved by specifying r > 0 to compute
log(x^(1/2^r))*2^r, at the cost of performing r square roots.
The caller must provide sufficient guard bits.
"""
for i in xrange(r):
x = isqrt_fast(x<<prec)
one = MPZ_ONE << prec
v = ((x-one)<<prec)//(x+one)
sign = v < 0
if sign:
v = -v
v2 = (v*v) >> prec
v4 = (v2*v2) >> prec
s0 = v
s1 = v//3
v = (v*v4) >> prec
k = 5
while v:
s0 += v // k
k += 2
s1 += v // k
v = (v*v4) >> prec
k += 2
s1 = (s1*v2) >> prec
s = (s0+s1) << (1+r)
if sign:
return -s
return s
def log_taylor_cached(x, prec):
"""
Fixed-point computation of log(x), assuming x in (0.5, 2)
and prec <= LOG_TAYLOR_PREC.
"""
n = x >> (prec-LOG_TAYLOR_SHIFT)
cached_prec = cache_prec_steps[prec]
dprec = cached_prec - prec
if (n, cached_prec) in log_taylor_cache:
a, log_a = log_taylor_cache[n, cached_prec]
else:
a = n << (cached_prec - LOG_TAYLOR_SHIFT)
log_a = log_taylor(a, cached_prec, 8)
log_taylor_cache[n, cached_prec] = (a, log_a)
a >>= dprec
log_a >>= dprec
u = ((x - a) << prec) // a
v = (u << prec) // ((MPZ_TWO << prec) + u)
v2 = (v*v) >> prec
v4 = (v2*v2) >> prec
s0 = v
s1 = v//3
v = (v*v4) >> prec
k = 5
while v:
s0 += v//k
k += 2
s1 += v//k
v = (v*v4) >> prec
k += 2
s1 = (s1*v2) >> prec
s = (s0+s1) << 1
return log_a + s
def mpf_log(x, prec, rnd=round_fast):
"""
Compute the natural logarithm of the mpf value x. If x is negative,
ComplexResult is raised.
"""
sign, man, exp, bc = x
#------------------------------------------------------------------
# Handle special values
if not man:
if x == fzero: return fninf
if x == finf: return finf
if x == fnan: return fnan
if sign:
raise ComplexResult("logarithm of a negative number")
wp = prec + 20
#------------------------------------------------------------------
# Handle log(2^n) = log(n)*2.
# Here we catch the only possible exact value, log(1) = 0
if man == 1:
if not exp:
return fzero
return from_man_exp(exp*ln2_fixed(wp), -wp, prec, rnd)
mag = exp+bc
abs_mag = abs(mag)
#------------------------------------------------------------------
# Handle x = 1+eps, where log(x) ~ x. We need to check for
# cancellation when moving to fixed-point math and compensate
# by increasing the precision. Note that abs_mag in (0, 1) <=>
# 0.5 < x < 2 and x != 1
if abs_mag <= 1:
# Calculate t = x-1 to measure distance from 1 in bits
tsign = 1-abs_mag
if tsign:
tman = (MPZ_ONE<<bc) - man
else:
tman = man - (MPZ_ONE<<(bc-1))
tbc = bitcount(tman)
cancellation = bc - tbc
if cancellation > wp:
t = normalize(tsign, tman, abs_mag-bc, tbc, tbc, 'n')
return mpf_perturb(t, tsign, prec, rnd)
else:
wp += cancellation
# TODO: if close enough to 1, we could use Taylor series
# even in the AGM precision range, since the Taylor series
# converges rapidly
#------------------------------------------------------------------
# Another special case:
# n*log(2) is a good enough approximation
if abs_mag > 10000:
if bitcount(abs_mag) > wp:
return from_man_exp(exp*ln2_fixed(wp), -wp, prec, rnd)
#------------------------------------------------------------------
# General case.
# Perform argument reduction using log(x) = log(x*2^n) - n*log(2):
# If we are in the Taylor precision range, choose magnitude 0 or 1.
# If we are in the AGM precision range, choose magnitude -m for
# some large m; benchmarking on one machine showed m = prec/20 to be
# optimal between 1000 and 100,000 digits.
if wp <= LOG_TAYLOR_PREC:
m = log_taylor_cached(lshift(man, wp-bc), wp)
if mag:
m += mag*ln2_fixed(wp)
else:
optimal_mag = -wp//LOG_AGM_MAG_PREC_RATIO
n = optimal_mag - mag
x = mpf_shift(x, n)
wp += (-optimal_mag)
m = -log_agm(to_fixed(x, wp), wp)
m -= n*ln2_fixed(wp)
return from_man_exp(m, -wp, prec, rnd)
def mpf_log_hypot(a, b, prec, rnd):
"""
Computes log(sqrt(a^2+b^2)) accurately.
"""
# If either a or b is inf/nan/0, assume it to be a
if not b[1]:
a, b = b, a
# a is inf/nan/0
if not a[1]:
# both are inf/nan/0
if not b[1]:
if a == b == fzero:
return fninf
if fnan in (a, b):
return fnan
# at least one term is (+/- inf)^2
return finf
# only a is inf/nan/0
if a == fzero:
# log(sqrt(0+b^2)) = log(|b|)
return mpf_log(mpf_abs(b), prec, rnd)
if a == fnan:
return fnan
return finf
# Exact
a2 = mpf_mul(a,a)
b2 = mpf_mul(b,b)
extra = 20
# Not exact
h2 = mpf_add(a2, b2, prec+extra)
cancelled = mpf_add(h2, fnone, 10)
mag_cancelled = cancelled[2]+cancelled[3]
# Just redo the sum exactly if necessary (could be smarter
# and avoid memory allocation when a or b is precisely 1
# and the other is tiny...)
if cancelled == fzero or mag_cancelled < -extra//2:
h2 = mpf_add(a2, b2, prec+extra-min(a2[2],b2[2]))
return mpf_shift(mpf_log(h2, prec, rnd), -1)
#----------------------------------------------------------------------------#
# #
# Exponential function #
# #
#----------------------------------------------------------------------------#
# The exponential function has a rapidly convergent Maclaurin series:
#
# exp(x) = 1 + x + x**2/2! + x**3/3! + x**4/4! + ...
#
# The series can be summed very easily using fixed-point arithmetic.
# The convergence can be improved further, using a trick due to
# Richard P. Brent: instead of computing exp(x) directly, we choose a
# small integer r (say, r=10) and compute exp(x/2**r)**(2**r).
# The optimal value for r depends on the Python platform, the magnitude
# of x and the target precision, and has to be estimated from
# experimental timings. One test with x ~= 0.3 showed that
# r = 2.2*prec**0.42 gave a good fit to the optimal values for r for
# prec between 1 and 10000 bits, on one particular machine.
# This optimization makes the summation about twice as fast at
# low precision levels and much faster at high precision
# (roughly five times faster at 1000 decimal digits).
# If |x| is very large, we first rewrite it as t + n*log(2) with the
# integer n chosen such that |t| <= log(2), and then calculate
# exp(x) as exp(t)*(2**n), using the Maclaurin series for exp(t)
# (the multiplication by 2**n just amounts to shifting the exponent).
# For very high precision use the newton method to compute exp from
# log_agm; for |x| very large or very small use
# exp(x + m) = exp(x) * e**m, m = int(n * math.log(2))
# Input: x * 2**prec
# Output: exp(x) * 2**(prec + r)
def exp_series(x, prec, r):
x >>= r
# 1 + x + x^2/2! + x^3/3! + x^4/4! + ... =
# (1 + x^2/2! + ...) + x * (1 + x^2/3! + ...)
s0 = s1 = (MPZ_ONE << prec)
k = 2
a = x2 = (x * x) >> prec
while a:
a = a // k
s0 += a
k += 1
a = a // k
s1 += a
a = (a * x2) >> prec
k += 1
# Calculate s**(2**r) by repeated squaring
s1 = (s1 * x) >> prec
s = s0 + s1
while r:
s = (s*s) >> prec
r -= 1
return s
def exp_series2(x, prec, r):
x >>= r
sign = 0
if x < 0:
sign = 1
x = -x
x2 = (x*x) >> prec
if prec < 1500:
s1 = a = x
k = 3
while a:
a = ((a * x2) >> prec) // (k*(k-1))
s1 += a
k += 2
else:
# use Smith's method:
# reduce the number of multiplication summing concurrently J series
# J=4
# Sinh(x) =
# (x + x^9/9! + ...) + x^2 * (x/3! + x^9/11! + ...) +
# x^4 * (x/5! + x^9/13! + ...) + x^6 * (x/7! + x^9/15! + ...)
J = 4
ax = [MPZ_ONE << prec, x2]
px = x2
asum = [x, x//6]
fact = 6
k = 4
for j in range(2, J):
px = (px * x2) >> prec
ax.append(px)
fact *= k*(k+1)
asum.append(x//fact)
k += 2
lx = (ax[-1]*x2) >> prec
p = asum[-1]
while p:
p = (p * lx) >> prec
for j in range(J):
p = p//(k*(k+1))
asum[j] += p
k += 2
s1 = 0
for i in range(1, J):
s1 += ax[i]*asum[i]
s1 = asum[0] + (s1 >> prec)
c1 = isqrt_fast((s1*s1) + (MPZ_ONE<<(2*prec)))
if sign:
s = c1 - s1
else:
s = c1 + s1
# Calculate s**(2**r) by repeated squaring
while r:
s = (s*s) >> prec
r -= 1
return s
# use the fourth order newton method, with step
# r = r + r * (h + h^2/2 + h^3/6 + h$/24)
# at each step the precision is quadrupled.
def exp_newton(x, prec):
extra = 10
r = mpf_exp(x, 60)
start = 50
prevp = start
for p in giant_steps(start, prec+extra, 4):
h = mpf_sub(x, mpf_log(r, p), p)
h2 = mpf_mul(h, h, p)
h3 = mpf_mul(h2, h, p)
h4 = mpf_mul(h2, h2, p)
t = mpf_add(h, mpf_shift(h2, -1), p)
t = mpf_add(t, mpf_div(h3, from_int(6, p), p), p)
t = mpf_add(t, mpf_div(h4, from_int(24, p), p), p)
t = mpf_mul(r, t, p)
r = mpf_add(r, t, p)
return r
# for precision larger than this limit, for x > 1, use the newton method
LIM_EXP_SERIES2 = 10000
# when the newton method is used, if x has mag=exp+bc larger than LIM_MAG
# shift it
LIM_MAG = 5
# table of values to determine if exp_series2 or exp_newton is faster,
# determined with benchmarking on a PC, with gmpy
ns_exp = [8,9,10,11,12,13,33,66,83,99,132,166,199,232,265,298,332,664]
precs_exp = [43000, 63000, 64000, 64000, 65000, 66000, 72000, 82000, 99000,
115000, 148000, 190000, 218000, 307000, 363000, 528000, 594000, 1650000]
def mpf_exp(x, prec, rnd=round_fast):
sign, man, exp, bc = x
if not man:
if not exp:
return fone
if x == fninf:
return fzero
return x
mag = bc+exp
# Fast handling e**n. TODO: the best cutoff depends on both the
# size of n and the precision.
if prec > 600 and exp >= 0:
e = mpf_e(prec+10+int(1.45*mag))
return mpf_pow_int(e, (-1)**sign *(man<<exp), prec, rnd)
if mag < -prec-10:
return mpf_perturb(fone, sign, prec, rnd)
# extra precision needs to be similar in magnitude to log_2(|x|)
# for the modulo reduction, plus r for the error from squaring r times
wp = prec + max(0, mag)
if wp < 300:
r = int(2*wp**0.4)
if mag < 0:
r = max(1, r + mag)
wp += r + 20
t = to_fixed(x, wp)
# abs(x) > 1?
if mag > 1:
lg2 = ln2_fixed(wp)
n, t = divmod(t, lg2)
else:
n = 0
man = exp_series(t, wp, r)
else:
use_newton = False
# put a bound on exp to avoid infinite recursion in exp_newton
# TODO find a good bound
if wp > LIM_EXP_SERIES2 and exp < 1000:
if mag > 0:
use_newton = True
elif mag <= 0 and -mag <= ns_exp[-1]:
i = bisect(ns_exp, -mag-1)
if i < len(ns_exp):
wp0 = precs_exp[i]
if wp > wp0:
use_newton = True
if not use_newton:
r = int(0.7 * wp**0.5)
if mag < 0:
r = max(1, r + mag)
wp += r + 20
t = to_fixed(x, wp)
if mag > 1:
lg2 = ln2_fixed(wp)
n, t = divmod(t, lg2)
else:
n = 0
man = exp_series2(t, wp, r)
else:
# if x is very small or very large use
# exp(x + m) = exp(x) * e**m
if mag > LIM_MAG:
wp += mag*10 + 100
n = int(mag * math.log(2)) + 1
x = mpf_sub(x, from_int(n, wp), wp)
elif mag <= 0:
wp += -mag*10 + 100
if mag < 0:
n = int(-mag * math.log(2)) + 1
x = mpf_add(x, from_int(n, wp), wp)
res = exp_newton(x, wp)
sign, man, exp, bc = res
if mag < 0:
t = mpf_pow_int(mpf_e(wp), n, wp)
res = mpf_div(res, t, wp)
sign, man, exp, bc = res
if mag > LIM_MAG:
t = mpf_pow_int(mpf_e(wp), n, wp)
res = mpf_mul(res, t, wp)
sign, man, exp, bc = res
return normalize(sign, man, exp, bc, prec, rnd)
bc = bitcount(man)
return normalize(0, man, int(-wp+n), bc, prec, rnd)
#----------------------------------------------------------------------------#
# #
# Trigonometric functions #
# #
#----------------------------------------------------------------------------#
def sin_taylor(x, prec):
x = MPZ(x)
x2 = (x*x) >> prec
s = a = x
k = 3
while a:
a = ((a * x2) >> prec) // (k*(1-k))
s += a
k += 2
return s
def cos_taylor(x, prec):
x = MPZ(x)
x2 = (x*x) >> prec
a = c = (MPZ_ONE<<prec)
k = 2
while a:
a = ((a * x2) >> prec) // (k*(1-k))
c += a
k += 2
return c
# Input: x * 2**prec
# Output: c * 2**(prec + r), s * 2**(prec + r)
def expi_series(x, prec, r):
x >>= r
one = MPZ_ONE << prec
x2 = (x*x) >> prec
s = x
a = x
k = 2
while a:
a = ((a * x2) >> prec) // (-k*(k+1))
s += a
k += 2
c = isqrt_fast((MPZ_ONE<<(2*prec)) - (s*s))
# Calculate (c + j*s)**(2**r) by repeated squaring
for j in range(r):
c, s = (c*c-s*s) >> prec, (2*c*s ) >> prec
return c, s
def reduce_angle(x, prec):
"""
Let x be a nonzero, finite mpf value defining angle (measured in
radians). Then reduce_trig(x, prec) returns (y, swaps, n) where:
y = (man, wp) is the reduced angle as a scaled fixed-point
number with precision wp, i.e. a floating-point number with
exponent -wp. The mantissa is positive and has width ~equal
to the input prec.
swaps = (swap_cos_sin, cos_sign, sin_sign)
Flags indicating the swaps that need to be applied
to (cos(y), sin(y)) to obtain (cos(x), sin(x))
n is an integer giving the original quadrant of x
Calculation of the quadrant
===========================
The integer n indices the quadrant of x. That is:
...
-pi < x < -pi/2 n = -2
-pi/2 < x < 0 n = -1
0 < x < pi/2 n = 0
pi/2 < x < pi n = 1
pi < x < 3*pi/2 n = 2
3*pi/2 < x < 2*pi n = 3
2*pi < x < 5*pi/2 n = 4
...
Note that n does not wrap around. A quadrant index normalized to
lie in [0, 1, 2, 3] can be found easily later on by computing
n % 4. Keeping the extended information in n is crucial for
interval arithmetic, as it allows one to distinguish between
whether two points of a sine wave lie next to each other on
a monotonic segment or are actually separated by a full
period (or several periods).
Note also that because is x is guaranteed to be rational, and
all roots of the sine/cosine are irrational, all inequalities are
strict. That is, we can always compute the correct quadrant.
Care is required to do ensure that this is done right.
Swaps
=====
The number y is a reduction of x to the first quadrant. This is
essentially x mod pi/2. In fact, we reduce y further, to the first
octant, by computing pi/2-x if x > pi/4.
Due to the translation and mirror symmetries of trigonometric
functions, this allows us to compute sin(x) or cos(x) by computing
+/-sin(y) or +/-cos(y). The point, of course, is that if x
is large, the Taylor series for y converges much more quickly
than the one for x.
"""
sign, man, exp, bc = x
magnitude = exp + bc
if not man:
return (0, 0), (0, 0, 0), 0
# Here we have abs(x) < 0.5. In this case no reduction is necessary.
# TODO: could also handle abs(x) < 1
if magnitude < 0:
# Quadrant is 0 or -1
n = -sign
swaps = (0, 0, sign)
fixed_exp = exp + bc - prec
delta = fixed_exp - exp
if delta < 0:
man <<= (-delta)
elif delta > 0:
man >>= delta
y = (man, -fixed_exp)
return y, swaps, n
i = 0
while 1:
cancellation_prec = 20 * 2**i
wp = prec + abs(magnitude) + cancellation_prec
pi1 = pi_fixed(wp)
pi2 = pi1 >> 1
pi4 = pi1 >> 2
# Find nearest multiple
n, y = divmod(to_fixed(x, wp), pi2)
# Interchange cos/sin ?
if y > pi4:
swap_cos_sin = 1
y = pi2 - y
else:
swap_cos_sin = 0
# Now, the catch is that x might be extremely close to a
# multiple of pi/2. This means accuracy is lost, and we may
# even end up in the wrong quadrant, which is bad news
# for interval arithmetic. This effect manifests by the
# fixed-point value of y becoming small. This is easy to check for.
if y >> (prec + magnitude - 10):
n = int(n)
swaps = swap_table[swap_cos_sin^(n%2)][n%4]
return (y>>magnitude, wp-magnitude), swaps, n
i += 1
swap_table = ((0,0,0),(0,1,0),(0,1,1),(0,0,1)), ((1,0,0),(1,1,0),(1,1,1),(1,0,1))
def calc_cos_sin(which, y, swaps, prec, cos_rnd, sin_rnd):
"""
Simultaneous computation of cos and sin (internal function).
"""
y, wp = y
swap_cos_sin, cos_sign, sin_sign = swaps
if swap_cos_sin:
which_compute = -which
else:
which_compute = which
# XXX: assumes no swaps
if not y:
return fone, fzero
# Tiny nonzero argument
if wp > prec*2 + 30:
y = from_man_exp(y, -wp)
if swap_cos_sin:
cos_rnd, sin_rnd = sin_rnd, cos_rnd
cos_sign, sin_sign = sin_sign, cos_sign
if cos_sign: cos = mpf_perturb(fnone, 0, prec, cos_rnd)
else: cos = mpf_perturb(fone, 1, prec, cos_rnd)
if sin_sign: sin = mpf_perturb(mpf_neg(y), 0, prec, sin_rnd)
else: sin = mpf_perturb(y, 1, prec, sin_rnd)
if swap_cos_sin:
cos, sin = sin, cos
return cos, sin
# Use standard Taylor series
if prec < 600:
if which_compute == 0:
sin = sin_taylor(y, wp)
# only need to evaluate one of the series
cos = isqrt_fast((MPZ_ONE<<(2*wp)) - sin*sin)
elif which_compute == 1:
sin = 0
cos = cos_taylor(y, wp)
elif which_compute == -1:
sin = sin_taylor(y, wp)
cos = 0
# Use exp(i*x) with Brent's trick
else:
r = int(0.137 * prec**0.579)
ep = r+20
cos, sin = expi_series(y<<ep, wp+ep, r)
cos >>= ep
sin >>= ep
if swap_cos_sin:
cos, sin = sin, cos
if cos_rnd is not round_nearest:
# Round and set correct signs
# XXX: this logic needs a second look
ONE = MPZ_ONE << wp
if cos_sign:
cos += (-1)**(cos_rnd in (round_ceiling, round_down))
cos = min(ONE, cos)
else:
cos += (-1)**(cos_rnd in (round_ceiling, round_up))
cos = min(ONE, cos)
if sin_sign:
sin += (-1)**(sin_rnd in (round_ceiling, round_down))
sin = min(ONE, sin)
else:
sin += (-1)**(sin_rnd in (round_ceiling, round_up))
sin = min(ONE, sin)
if which != -1:
cos = normalize(cos_sign, cos, -wp, bitcount(cos), prec, cos_rnd)
if which != 1:
sin = normalize(sin_sign, sin, -wp, bitcount(sin), prec, sin_rnd)
return cos, sin
def mpf_cos_sin(x, prec, rnd=round_fast, which=0):
"""
Computes (cos(x), sin(x)). The parameter 'which' can disable
evaluation of either cos or sin:
0 -- return (cos(x), sin(x), n)
1 -- return (cos(x), -, n)
-1 -- return (-, sin(x), n)
If only one function is wanted, this is slightly
faster at low precision.
"""
sign, man, exp, bc = x
# Exact (or special) cases
if not man:
if exp:
return (fnan, fnan)
else:
return (fone, fzero)
y, swaps, n = reduce_angle(x, prec+10)
return calc_cos_sin(which, y, swaps, prec, rnd, rnd)
def mpf_cos(x, prec, rnd=round_fast):
return mpf_cos_sin(x, prec, rnd, 1)[0]
def mpf_sin(x, prec, rnd=round_fast):
return mpf_cos_sin(x, prec, rnd, -1)[1]
def mpf_tan(x, prec, rnd=round_fast):
c, s = mpf_cos_sin(x, prec+20)
return mpf_div(s, c, prec, rnd)
# Accurate computation of cos(pi*x) and sin(pi*x) is needed by
# reflection formulas for gamma, polygamma, zeta, etc
def mpf_cos_sin_pi(x, prec, rnd=round_fast):
"""Accurate computation of (cos(pi*x), sin(pi*x))
for x close to an integer"""
sign, man, exp, bc = x
if not man:
return mpf_cos_sin(x, prec, rnd)
# Exactly an integer or half-integer?
if exp >= -1:
if exp == -1:
c = fzero
s = (fone, fnone)[bool(man & 2) ^ sign]
elif exp == 0:
c, s = (fnone, fzero)
else:
c, s = (fone, fzero)
return c, s
# Close to 0 ?
size = exp + bc
if size < -(prec+5):
c = mpf_perturb(fone, 1, prec, rnd)
s = mpf_perturb(mpf_mul(x, mpf_pi(prec)), sign, prec, rnd)
return c, s
if sign:
man = -man
# Subtract nearest half-integer (= modulo pi/2)
nhint = ((man >> (-exp-2)) + 1) >> 1
man = man - (nhint << (-exp-1))
x = from_man_exp(man, exp, prec)
x = mpf_mul(x, mpf_pi(prec), prec)
# XXX: with some more work, could call calc_cos_sin,
# to save some time and to get rounding right
case = nhint % 4
if case == 0:
c, s = mpf_cos_sin(x, prec, rnd)
elif case == 1:
s, c = mpf_cos_sin(x, prec, rnd)
c = mpf_neg(c)
elif case == 2:
c, s = mpf_cos_sin(x, prec, rnd)
c = mpf_neg(c)
s = mpf_neg(s)
else:
s, c = mpf_cos_sin(x, prec, rnd)
s = mpf_neg(s)
return c, s
def mpf_cos_pi(x, prec, rnd=round_fast):
return mpf_cos_sin_pi(x, prec, rnd)[0]
def mpf_sin_pi(x, prec, rnd=round_fast):
return mpf_cos_sin_pi(x, prec, rnd)[1]
#----------------------------------------------------------------------
# Hyperbolic functions
#
def sinh_taylor(x, prec):
x = MPZ(x)
x2 = (x*x) >> prec
s = a = x
k = 3
while a:
a = ((a * x2) >> prec) // (k*(k-1))
s += a
k += 2
return s
def mpf_cosh_sinh(x, prec, rnd=round_fast, tanh=0):
"""Simultaneously compute (cosh(x), sinh(x)) for real x"""
sign, man, exp, bc = x
if (not man) and exp:
if tanh:
if x == finf: return fone
if x == fninf: return fnone
return fnan
if x == finf: return (finf, finf)
if x == fninf: return (finf, fninf)
return fnan, fnan
if sign:
man = -man
mag = exp + bc
prec2 = prec + 20
if mag < -3:
# Extremely close to 0, sinh(x) ~= x and cosh(x) ~= 1
if mag < -prec-2:
if tanh:
return mpf_perturb(x, 1-sign, prec, rnd)
cosh = mpf_perturb(fone, 0, prec, rnd)
sinh = mpf_perturb(x, sign, prec, rnd)
return cosh, sinh
# Avoid cancellation when computing sinh
# TODO: might be faster to use sinh series directly
prec2 += (-mag) + 4
# In the general case, we use
# cosh(x) = (exp(x) + exp(-x))/2
# sinh(x) = (exp(x) - exp(-x))/2
# and note that the exponential only needs to be computed once.
ep = mpf_exp(x, prec2)
em = mpf_div(fone, ep, prec2)
if tanh:
ch = mpf_add(ep, em, prec2, rnd)
sh = mpf_sub(ep, em, prec2, rnd)
return mpf_div(sh, ch, prec, rnd)
else:
ch = mpf_shift(mpf_add(ep, em, prec, rnd), -1)
sh = mpf_shift(mpf_sub(ep, em, prec, rnd), -1)
return ch, sh
def mpf_cosh(x, prec, rnd=round_fast):
"""Compute cosh(x) for a real argument x"""
return mpf_cosh_sinh(x, prec, rnd)[0]
def mpf_sinh(x, prec, rnd=round_fast):
"""Compute sinh(x) for a real argument x"""
return mpf_cosh_sinh(x, prec, rnd)[1]
def mpf_tanh(x, prec, rnd=round_fast):
"""Compute tanh(x) for a real argument x"""
return mpf_cosh_sinh(x, prec, rnd, tanh=1)
#----------------------------------------------------------------------
# Inverse tangent
#
def atan_newton(x, prec):
if prec >= 100:
r = math.atan((x>>(prec-53))/2.0**53)
else:
r = math.atan(x/2.0**prec)
prevp = 50
r = int(r * 2.0**53) >> (53-prevp)
extra_p = 100
for p in giant_steps(prevp, prec):
s = int(0.137 * p**0.579)
p += s + 50
r = r << (p-prevp)
cos, sin = expi_series(r, p, s)
tan = (sin << p) // cos
a = ((tan - rshift(x, prec-p)) << p) // ((MPZ_ONE<<p) + ((tan**2)>>p))
r = r - a
prevp = p
return rshift(r, prevp-prec)
ATAN_TAYLOR_PREC = 3000
ATAN_TAYLOR_SHIFT = 7 # steps of size 2^-N
atan_taylor_cache = {}
def atan_taylor_get_cached(n, prec):
# Taylor series with caching wins up to huge precisions
# To avoid unnecessary precomputation at low precision, we
# do it in steps
# Round to next power of 2
prec2 = (1<<(bitcount(prec-1))) + 20
dprec = prec2 - prec
if (n, prec2) in atan_taylor_cache:
a, atan_a = atan_taylor_cache[n, prec2]
else:
a = n << (prec2 - ATAN_TAYLOR_SHIFT)
atan_a = atan_newton(a, prec2)
atan_taylor_cache[n, prec2] = (a, atan_a)
return (a >> dprec), (atan_a >> dprec)
def atan_taylor(x, prec):
n = (x >> (prec-ATAN_TAYLOR_SHIFT))
a, atan_a = atan_taylor_get_cached(n, prec)
d = x - a
s0 = v = (d << prec) // ((a**2 >> prec) + (a*d >> prec) + (MPZ_ONE << prec))
v2 = (v**2 >> prec)
v4 = (v2 * v2) >> prec
s1 = v//3
v = (v * v4) >> prec
k = 5
while v:
s0 += v // k
k += 2
s1 += v // k
v = (v * v4) >> prec
k += 2
s1 = (s1 * v2) >> prec
s = s0 - s1
return atan_a + s
def atan_inf(sign, prec, rnd):
if not sign:
return mpf_shift(mpf_pi(prec, rnd), -1)
return mpf_neg(mpf_shift(mpf_pi(prec, negative_rnd[rnd]), -1))
def mpf_atan(x, prec, rnd=round_fast):
sign, man, exp, bc = x
if not man:
if x == fzero: return fzero
if x == finf: return atan_inf(0, prec, rnd)
if x == fninf: return atan_inf(1, prec, rnd)
return fnan
mag = exp + bc
# Essentially infinity
if mag > prec+20:
return atan_inf(sign, prec, rnd)
# Essentially ~ x
if -mag > prec+20:
return mpf_perturb(x, 1-sign, prec, rnd)
wp = prec + 30 + abs(mag)
# For large x, use atan(x) = pi/2 - atan(1/x)
if mag >= 2:
x = mpf_rdiv_int(1, x, wp)
reciprocal = True
else:
reciprocal = False
t = to_fixed(x, wp)
if sign:
t = -t
if wp < ATAN_TAYLOR_PREC:
a = atan_taylor(t, wp)
else:
a = atan_newton(t, wp)
if reciprocal:
a = ((pi_fixed(wp)>>1)+1) - a
if sign:
a = -a
return from_man_exp(a, -wp, prec, rnd)
# TODO: cleanup the special cases
def mpf_atan2(y, x, prec, rnd=round_fast):
xsign, xman, xexp, xbc = x
ysign, yman, yexp, ybc = y
if not yman:
if y == fzero and x != fnan:
if mpf_sign(x) >= 0:
return fzero
return mpf_pi(prec, rnd)
if y in (finf, fninf):
if x in (finf, fninf):
return fnan
# pi/2
if y == finf:
return mpf_shift(mpf_pi(prec, rnd), -1)
# -pi/2
return mpf_neg(mpf_shift(mpf_pi(prec, negative_rnd[rnd]), -1))
return fnan
if ysign:
return mpf_neg(mpf_atan2(mpf_neg(y), x, prec, negative_rnd[rnd]))
if not xman:
if x == fnan:
return fnan
if x == finf:
return fzero
if x == fninf:
return mpf_pi(prec, rnd)
if y == fzero:
return fzero
return mpf_shift(mpf_pi(prec, rnd), -1)
tquo = mpf_atan(mpf_div(y, x, prec+4), prec+4)
if xsign:
return mpf_add(mpf_pi(prec+4), tquo, prec, rnd)
else:
return mpf_pos(tquo, prec, rnd)
def mpf_asin(x, prec, rnd=round_fast):
sign, man, exp, bc = x
if bc+exp > 0 and x not in (fone, fnone):
raise ComplexResult("asin(x) is real only for -1 <= x <= 1")
# asin(x) = 2*atan(x/(1+sqrt(1-x**2)))
wp = prec + 15
a = mpf_mul(x, x)
b = mpf_add(fone, mpf_sqrt(mpf_sub(fone, a, wp), wp), wp)
c = mpf_div(x, b, wp)
return mpf_shift(mpf_atan(c, prec, rnd), 1)
def mpf_acos(x, prec, rnd=round_fast):
# acos(x) = 2*atan(sqrt(1-x**2)/(1+x))
sign, man, exp, bc = x
if bc + exp > 0:
if x not in (fone, fnone):
raise ComplexResult("acos(x) is real only for -1 <= x <= 1")
if x == fnone:
return mpf_pi(prec, rnd)
wp = prec + 15
a = mpf_mul(x, x)
b = mpf_sqrt(mpf_sub(fone, a, wp), wp)
c = mpf_div(b, mpf_add(fone, x, wp), wp)
return mpf_shift(mpf_atan(c, prec, rnd), 1)
def mpf_asinh(x, prec, rnd=round_fast):
wp = prec + 20
sign, man, exp, bc = x
mag = exp+bc
if mag < -8:
if mag < -wp:
return mpf_perturb(x, 1-sign, prec, rnd)
wp += (-mag)
# asinh(x) = log(x+sqrt(x**2+1))
# use reflection symmetry to avoid cancellation
q = mpf_sqrt(mpf_add(mpf_mul(x, x), fone, wp), wp)
q = mpf_add(mpf_abs(x), q, wp)
if sign:
return mpf_neg(mpf_log(q, prec, negative_rnd[rnd]))
else:
return mpf_log(q, prec, rnd)
def mpf_acosh(x, prec, rnd=round_fast):
# acosh(x) = log(x+sqrt(x**2-1))
wp = prec + 15
if mpf_cmp(x, fone) == -1:
raise ComplexResult("acosh(x) is real only for x >= 1")
q = mpf_sqrt(mpf_add(mpf_mul(x,x), fnone, wp), wp)
return mpf_log(mpf_add(x, q, wp), prec, rnd)
def mpf_atanh(x, prec, rnd=round_fast):
# atanh(x) = log((1+x)/(1-x))/2
sign, man, exp, bc = x
if (not man) and exp:
if x in (fzero, fnan):
return x
raise ComplexResult("atanh(x) is real only for -1 <= x <= 1")
mag = bc + exp
if mag > 0:
if mag == 1 and man == 1:
return [finf, fninf][sign]
raise ComplexResult("atanh(x) is real only for -1 <= x <= 1")
wp = prec + 15
if mag < -8:
if mag < -wp:
return mpf_perturb(x, sign, prec, rnd)
wp += (-mag)
a = mpf_add(x, fone, wp)
b = mpf_sub(fone, x, wp)
return mpf_shift(mpf_log(mpf_div(a, b, wp), prec, rnd), -1)
def mpf_fibonacci(x, prec, rnd=round_fast):
sign, man, exp, bc = x
if not man:
if x == fninf:
return fnan
return x
# F(2^n) ~= 2^(2^n)
size = abs(exp+bc)
if exp >= 0:
# Exact
if size < 10 or size <= bitcount(prec):
return from_int(ifib(to_int(x)), prec, rnd)
# Use the modified Binet formula
wp = prec + size + 20
a = mpf_phi(wp)
b = mpf_add(mpf_shift(a, 1), fnone, wp)
u = mpf_pow(a, x, wp)
v = mpf_cos_pi(x, wp)
v = mpf_div(v, u, wp)
u = mpf_sub(u, v, wp)
u = mpf_div(u, b, prec, rnd)
return u