def __exit__(self, typ, value, tb):
"""
EXAMPLE::
sage: from sage.libs.giac import GiacSettingsDefaultContext # optional - giacpy
sage: from giacpy import giacsettings # optional - giacpy
sage: giacsettings.proba_epsilon = 1e-16 # optional - giacpy
sage: with GiacSettingsDefaultContext(): giacsettings.proba_epsilon = 1e-30 # optional - giacpy
sage: giacsettings.proba_epsilon < 1e-20 # optional - giacpy
False
"""
try:
from giacpy import giacsettings, libgiac
except ImportError:
raise ImportError("""One of the optional packages giac or giacpy is missing""")
# Restore the debug level first to not have messages at each modification
libgiac('debug_infolevel')(self.debuginfolevel)
# NB: giacsettings.epsilon has a different meaning that giacsettings.proba_epsilon.
giacsettings.proba_epsilon = self.proba_epsilon
giacsettings.threads = self.threads
def __enter__(self):
"""
EXAMPLE::
sage: from sage.libs.giac import GiacSettingsDefaultContext # optional - giacpy
sage: from giacpy import giacsettings # optional - giacpy
sage: giacsettings.proba_epsilon = 1e-16 # optional - giacpy
sage: with GiacSettingsDefaultContext(): giacsettings.proba_epsilon = 1e-12 # optional - giacpy
sage: giacsettings.proba_epsilon < 1e-14 # optional - giacpy
True
"""
try:
from giacpy import giacsettings, libgiac
except ImportError:
raise ImportError("""One of the optional packages giac or giacpy is missing""")
self.proba_epsilon = giacsettings.proba_epsilon
self.threads = giacsettings.threads
# Change the debug level at the end to not have messages at each modification
self.debuginfolevel = libgiac('debug_infolevel()')
def groebner_basis(gens, proba_epsilon=None, threads=None, prot=False, *args, **kwds):
"""
Computes a Groebner Basis of an ideal using giacpy. The result is
automatically converted to sage.
INPUT:
- ``gens`` - an ideal (or a list) of polynomials over a prime field
of characteristic 0 or p<2^31
- ``proba_epsilon`` - (default: None) majoration of the probability
of a wrong answer when probabilistic algorithms are allowed.
* if ``proba_epsilon`` is None, the value of
``sage.structure.proof.all.polynomial()`` is taken. If it is
false then the global ``giacpy.giacsettings.proba_epsilon`` is
used.
* if ``proba_epsilon`` is 0, probabilistic algorithms are
disabled.
- ``threads`` - (default: None) Maximal number of threads allowed
for giac. If None, the global ``giacpy.giacsettings.threads`` is
considered.
- ``prot`` - (default: False) if True print detailled informations
OUTPUT:
Polynomial sequence of the reduced Groebner basis.
EXAMPLES::
sage: from sage.libs.giac import groebner_basis as gb_giac # optional - giacpy
sage: P = PolynomialRing(GF(previous_prime(2**31)), 6, 'x') # optional - giacpy
sage: I = sage.rings.ideal.Cyclic(P) # optional - giacpy
sage: B=gb_giac(I.gens());B # optional - giacpy
<BLANKLINE>
// Groebner basis computation time ...
Polynomial Sequence with 45 Polynomials in 6 Variables
sage: B.is_groebner() # optional - giacpy
True
Computations over QQ can benefit from
* a probabilistic lifting::
sage: P = PolynomialRing(QQ,5, 'x') # optional - giacpy
sage: I = ideal([P.random_element(3,7) for j in range(5)]) # optional - giacpy
sage: B1 = gb_giac(I.gens(),1e-16) # optional - giacpy, long time (1s)
Running a probabilistic check for the reconstructed Groebner basis.
If successfull, error probability is less than 1e-16 ...
sage: sage.structure.proof.all.polynomial(True) # optional - giacpy
sage: B2 = gb_giac(I.gens()) # optional - giacpy, long time (4s)
<BLANKLINE>
// Groebner basis computation time...
sage: B1==B2 # optional - giacpy, long time
True
sage: B1.is_groebner() # optional - giacpy, long time (20s)
True
* multi threaded operations::
sage: P = PolynomialRing(QQ, 8, 'x') # optional - giacpy
sage: I=sage.rings.ideal.Cyclic(P) # optional - giacpy
sage: time B = gb_giac(I.gens(),1e-6,threads=2) # doctest: +SKIP
Running a probabilistic check for the reconstructed Groebner basis...
Time: CPU 168.98 s, Wall: 94.13 s
You can get detailled information by setting ``prot=True``
::
sage: I=sage.rings.ideal.Katsura(P) # optional - giacpy
sage: gb_giac(I,prot=True) # optional - giacpy, random, long time (3s)
9381383 begin computing basis modulo 535718473
9381501 begin new iteration zmod, number of pairs: 8, base size: 8
...end, basis size 74 prime number 1
G=Vector [0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,...
...creating reconstruction #0
...
++++++++basis size 74
checking pairs for i=0, j=
checking pairs for i=1, j=2,6,12,17,19,24,29,34,39,42,43,48,56,61,64,69,
...
checking pairs for i=72, j=73,
checking pairs for i=73, j=
Number of critical pairs to check 373
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++...
Successfull check of 373 critical pairs
12380865 end final check
Polynomial Sequence with 74 Polynomials in 8 Variables
TESTS::
sage: from giacpy import libgiac # optional - giacpy
sage: libgiac("x2:=22; x4:='whywouldyoudothis'") # optional - giacpy
22,whywouldyoudothis
sage: gb_giac(I) # optional - giacpy
Traceback (most recent call last):
...
ValueError: Variables names ['x2', 'x4'] conflict in giac. Change them or purge them from in giac with libgiac.purge('x2')
sage: libgiac.purge('x2'),libgiac.purge('x4') # optional - giacpy
(22, whywouldyoudothis)
sage: gb_giac(I) # optional - giacpy, long time (3s)
<BLANKLINE>
// Groebner basis computation time...
Polynomial Sequence with 74 Polynomials in 8 Variables
sage: I=ideal(P(0),P(0)) # optional - giacpy
sage: I.groebner_basis() == gb_giac(I) # optional - giacpy
True
"""
try:
from giacpy import libgiac, giacsettings
except ImportError:
raise ImportError("""One of the optional packages giac or giacpy is missing""")
try:
iter(gens)
except TypeError:
gens = gens.gens()
# get the ring from gens
P = next(iter(gens)).parent()
K = P.base_ring()
p = K.characteristic()
# check if the ideal is zero. (giac 1.2.0.19 segfault)
from sage.rings.ideal import Ideal
if (Ideal(gens)).is_zero():
return PolynomialSequence([P(0)], P, immutable=True)
# check for name confusions
blackgiacconstants = ['i', 'e'] # NB e^k is expanded to exp(k)
blacklist = blackgiacconstants + [str(j) for j in libgiac.VARS()]
problematicnames = list(set(P.gens_dict().keys()).intersection(blacklist))
if(len(problematicnames)>0):
raise ValueError("Variables names %s conflict in giac. Change them or purge them from in giac with libgiac.purge(\'%s\')"
%(problematicnames, problematicnames[0]))
if K.is_prime_field() and p == 0:
F = libgiac(gens)
elif K.is_prime_field() and p < 2**31:
F = (libgiac(gens) % p)
else:
raise NotImplementedError("Only prime fields of cardinal < 2^31 are implemented in Giac for Groebner bases.")
if P.term_order() != "degrevlex":
raise NotImplementedError("Only degrevlex term orderings are supported in Giac Groebner bases.")
# proof or probabilistic reconstruction
if proba_epsilon is None:
if proof_polynomial():
giacsettings.proba_epsilon = 0
else:
giacsettings.proba_epsilon = 1e-15
else:
giacsettings.proba_epsilon = proba_epsilon
# prot
if prot:
libgiac('debug_infolevel(2)')
# threads
if threads is not None:
giacsettings.threads = threads
# compute de groebner basis with giac
gb_giac = F.gbasis([P.gens()], "revlex")
return PolynomialSequence(gb_giac, P, immutable=True)
def rational_univariate_representation(ideal):
I = libgiac(ideal.gens())
vars = list(ideal.ring().gens())
return I.gbasis(libgiac(vars), 'rur')
