def numeric_discovery_measures(model):
nrules = model.number_rules
#nrows= len(model.target_values)
kl_supp, kl_usg, wkl_supp, wkl_usg, wkl_sum = np.zeros(nrules), np.zeros(
nrules), np.zeros(nrules), np.zeros(nrules), np.zeros(nrules)
wacc_supp, wacc_usg = np.zeros(nrules), np.zeros(nrules)
support, usage = np.zeros(nrules), np.zeros(nrules)
std_rules = [stat["variance"]**0.5 for stat in model.statistic_rules]
std_rulesalternative = []
tid_covered = mpz()
for r in range(nrules):
tid_support = model.bitset_rules[r]
tid_usage = model.bitset_rules[r] & ~tid_covered
tid_covered = tid_covered | tid_support
aux_bitset = xmpz(tid_support)
idx_bits = list(aux_bitset.iter_set())
values_support = model.target_values[idx_bits]
aux_bitset = xmpz(tid_usage)
idx_bits = list(aux_bitset.iter_set())
values_usage = model.target_values[idx_bits]
support[r] = len(values_support)
usage[r] = len(values_usage)
kl_supp[r], wkl_supp[r] = kullbackleibler_gaussian_paramters(
model, values_support)
kl_usg[r], wkl_usg[r] = kullbackleibler_gaussian_paramters(
model, values_usage)
std_rulesalternative.append(np.std(values_usage))
wacc_supp[r] = wracc_numeric(model, values_support)
wacc_usg[r] = wracc_numeric(model, values_usage)
#print(wkl_usg)
wkl_sum = sum(wkl_usg)
# Average them all!!!!
measures = dict()
measures["avg_supp"] = np.mean(support)
measures["kl_supp"] = np.mean(kl_supp)
measures["wkl_supp"] = np.mean(wkl_supp)
measures["avg_usg"] = np.mean(usage)
measures["kl_usg"] = np.mean(kl_usg)
measures["wkl_usg"] = np.mean(wkl_usg)
measures["wacc_supp"] = np.mean(wacc_supp)
measures["wacc_usg"] = np.mean(wacc_usg)
uptm = np.triu_indices(nrules - 1, 1)
measures["jacc_avg"] = np.sum(model.jaccard_matrix) / len(uptm[0])
measures["n_rules"] = model.number_rules
measures["avg_items"] = sum([len(ant) for ant in model.antecedent_raw
]) / model.number_rules
measures["wkl_sum"] = wkl_sum
measures["std_rules"] = np.mean(std_rules)
measures["top1_std"] = std_rules[0]
measures["length_orig"] = model.length_original
measures["length_final"] = model.length_data + model.length_model
measures["length_ratio"] = model.length_ratio
return measures
def nominal_discovery_measures(default_prob_per_class, subgroup_bitarray, X,
Y):
nrules = len(subgroup_bitarray)
nusage_fail = 0 # number of rules that the usage fails
nrows = X.shape[0]
data_prob_class = default_prob_per_class
wkl_supp, wkl_usg, wkl_sum = np.zeros(nrules), np.zeros(nrules), np.zeros(
nrules)
wacc_supp, wacc_usg = np.zeros(nrules), np.zeros(nrules)
support, usage = np.zeros(nrules), np.zeros(nrules)
tid_covered = mpz()
list_bitsets = []
number_targets = len(data_prob_class)
for r, bitarray in enumerate(subgroup_bitarray):
tid_support = bitarray
list_bitsets.append(tid_support)
tid_usage = tid_support & ~tid_covered
tid_covered = tid_covered | tid_support
aux_bitset = xmpz(tid_support)
idx_bits = list(aux_bitset.iter_set())
values_support = Y.iloc[idx_bits, :].values
aux_bitset = xmpz(tid_usage)
idx_bits = list(aux_bitset.iter_set())
values_usage = Y.iloc[idx_bits, :].values
support[r] = values_support.shape[0]
usage[r] = values_usage.shape[0]
wkl_supp[r], wacc_supp[r] = wkl_wracc(data_prob_class, values_support,
nrows, number_targets)
if usage[r] != 0:
wkl_usg[r], wacc_usg[r] = wkl_wracc(data_prob_class, values_usage,
nrows, number_targets)
else:
nusage_fail += 1
wkl_sum = sum(wkl_usg)
# Average them all!!!!
measures = dict()
measures["avg_supp"] = np.mean(support)
measures["wkl_supp"] = np.mean(wkl_supp)
measures["avg_usg"] = np.sum(usage) / (nrules - nusage_fail)
measures["wkl_usg"] = np.sum(wkl_usg) / (nrules - nusage_fail)
measures["wacc_supp"] = np.mean(wacc_supp)
measures["wacc_usg"] = np.sum(wacc_usg) / (nrules - nusage_fail)
measures["jacc_avg"], jaccard_matrix = jaccard_index_model(list_bitsets)
measures["n_rules"] = nrules - nusage_fail
#measures["avg_items"] = sum([len(sg.pattern) for sg in rulelist.subgroups]) / rulelist.number_rules
measures["wkl_sum"] = wkl_sum
measures["wkl_sum_norm"] = wkl_sum / X.shape[0]
measures["wacc_supp_sum"] = np.sum(wacc_supp)
measures["wacc_usg_sum"] = np.sum(wacc_usg)
return measures
def nominal_discovery_measures(rulelist, X, Y):
nrules = rulelist.number_rules
nrows = X.shape[0]
data_prob_class = rulelist.default_rule_statistics.prob_per_classes
wkl_supp, wkl_usg, wkl_sum = np.zeros(nrules), np.zeros(nrules), np.zeros(
nrules)
wacc_supp, wacc_usg = np.zeros(nrules), np.zeros(nrules)
support, usage = np.zeros(nrules), np.zeros(nrules)
tid_covered = mpz()
list_bitsets = []
number_targets = len(rulelist.default_rule_statistics.prob_per_classes)
for r in range(nrules):
tid_support = rulelist.subgroups[r].bitarray
list_bitsets.append(tid_support)
tid_usage = tid_support & ~tid_covered
tid_covered = tid_covered | tid_support
aux_bitset = xmpz(tid_support)
idx_bits = list(aux_bitset.iter_set())
values_support = Y.iloc[idx_bits, :].values
aux_bitset = xmpz(tid_usage)
idx_bits = list(aux_bitset.iter_set())
values_usage = Y.iloc[idx_bits, :].values
support[r] = values_support.shape[0]
usage[r] = values_usage.shape[0]
wkl_supp[r], wacc_supp[r] = wkl_wracc(data_prob_class, values_support,
nrows, number_targets)
wkl_usg[r], wacc_usg[r] = wkl_wracc(data_prob_class, values_usage,
nrows, number_targets)
wkl_sum = sum(wkl_usg)
# Average them all!!!!
measures = dict()
measures["avg_supp"] = np.mean(support)
measures["wkl_supp"] = np.mean(wkl_supp)
measures["avg_usg"] = np.mean(usage)
measures["wkl_usg"] = np.mean(wkl_usg)
measures["wacc_supp"] = np.mean(wacc_supp)
measures["wacc_usg"] = np.mean(wacc_usg)
measures["jacc_avg"], jaccard_matrix = jaccard_index_model(list_bitsets)
measures["n_rules"] = rulelist.number_rules
measures["avg_items"] = sum([len(sg.pattern) for sg in rulelist.subgroups
]) / rulelist.number_rules
measures["wkl_sum"] = wkl_sum
measures["wkl_sum_norm"] = wkl_sum / X.shape[0]
measures["wacc_supp_sum"] = np.sum(wacc_supp)
measures["wacc_usg_sum"] = np.sum(wacc_usg)
measures["length_orig"] = rulelist.length_original
measures["length_final"] = rulelist.length_data + rulelist.length_model
measures["length_ratio"] = rulelist.length_ratio
return measures
def countTime():
millis = 0
for n in mersenne:
millis = int(round(time.time() * 1000000000000))
p = gmpy2.xmpz(n)
s = gmpy2.xmpz(2)
# (s**p)-1
millis = int(round(time.time() * 1000000000000)) - millis
print "time: %d"%millis
with open('file.txt', 'a') as f:
f.write("%d\n"%(n))
with open('mersenne.txt', 'a') as f:
f.write("%d\n"%(millis))
def plotMillerTime(p): multiVar = 10000000000 mills = int(round(time.time()*multiVar)) n = (gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p))+(gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p)-1) while not miller_rabin.millerRabin(n, 2): n = (gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p))+(gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p)-1) mills = int(round(time.time()*multiVar)) - mills return (mills, bit_length(n), totalDigits(n))
def tetration_mod(a:int,height:int,digits:int)->int:
"""Gets the rightmost digits of a tetration operation
Notes:
1. Wikipedia's entry on [Graham's number](https://en.wikipedia.org/wiki/Graham%27s_number)
describes an algorithm for getting the rightmost digits of a tetration operation:
> A simple algorithm for computing these digits may be described as follows:
> let x = 3, then iterate, d times, the assignment x = 3**x mod 10d. Except for omitting
> any leading 0s, the final value assigned to x (as a base-ten numeral) is then composed
> of the d rightmost decimal digits of 3↑↑n, for all n > d. (If the final value of x
> has fewer than d digits, then the required number of leading 0s must be added.)
2. Native python is painfully slow; using [GMP](http://gmplib.org) for
integer operations is way faster.
See also https://gmpy2.readthedocs.io/en/latest/index.html
Args:
a (int): Argument
height (int): Height
digits (int): Number of rightmost digits to capture
Returns:
int: Rightmost digits
"""
result = gmpy2.xmpz(a)
for i in range(1,min(height,digits+1)):
result = (a**result)%(10**digits)
return result
def mk_bitarray(self, index, partition='all'):
"""Produces a bitarray representing the presence / absence of families in the organism using the provided index
The bitarray is stored in the :attr:`bitarray` attribute and is a :class:`gmpy2.xmpz` type.
:param index: The index computed by :func:`ppanggolin.pangenome.Pangenome.getIndex`
:type index: dict[:class:`ppanggolin.genome.Organism`, int]
"""
self.bitarray = gmpy2.xmpz() # pylint: disable=no-member
if partition == 'all':
logging.getLogger().debug(f"all")
for fam in self.families:
self.bitarray[index[fam]] = 1
elif partition == 'persistent':
logging.getLogger().debug(f"persistent")
for fam in self.families:
if fam.namedPartition in ['persistent']:
self.bitarray[index[fam]] = 1
elif partition in ['shell', 'cloud']:
logging.getLogger().debug(f"shell, cloud")
for fam in self.families:
if fam.namedPartition == partition:
self.bitarray[index[fam]] = 1
elif partition == 'accessory':
logging.getLogger().debug(f"accessory")
for fam in self.families:
if fam.namedPartition in ['shell', 'cloud']:
self.bitarray[index[fam]] = 1
else:
raise Exception(
"There is not any partition corresponding please report a github issue"
)
def P_generic(m, _x, debug=False):
"""
Used for finding s0, this is a generic implementation because m might be any value
and optimizations aren't needed anyway since this is called once per number.
Algorithm found in: https://vixra.org/pdf/1303.0195v1.pdf
Test data can ve verified using:
# for checking s0 for k=2001 b=2
https://www.wolframalpha.com/input/?i=2+*+chebyshevT%282*2001%2F2%2C+chebyshevT%282%2F2%2C+2%29%29
"""
print("M {} X {}".format(m, _x))
m = mpz(m)
x = mpz(_x)
a = mpfr(mpfr(2)**-m)
inner = pow(x, mpz(2))
inner -= mpz(4)
inner = sqrt(inner)
#inner = x - (sqrt4 / x) # potential replacement in cases x >= 5
x += inner
print("x before pow: {:.64f}".format(x))
print("a before pow: {:.64f}".format(a))
x **= m
x *= a
result = xmpz(round_away(x))
if debug:
print("P_gen: {} {}".format(m, _x))
return result
def generate_p_q(L, N):
g = N # g >= 160
n = (L - 1) // g #n es igual cociente entre 1023/160 = 6
b = (L - 1) % g #b es igual al residuo entre 1023/160 = 63
while True:
# generate q
while True:
s = xmpz(
randrange(1, 2**(g))
) #randrange es una funcion que brinda un numero aleatorio entre 1 y 2^160
a = sha1(to_binary(s)).hexdigest(
) #se convierte s a binario y se le saca el hash en hexadecimal
zz = xmpz(
(s + 1) % (2**g)
) #se ubica un numero + 1 a s y se confirma que este en el cuerpo de 2^160
z = sha1(to_binary(zz)).hexdigest(
) #se convierte zz a binario y se le sacah el hash en hexadecimal
U = int(a, 16) ^ int(
z, 16
) #se covierte los hashes a enteros en base 16 y se realiza una operacion XOR
mask = 2**(
N - 1
) + 1 #se genera una mascara que sea un bit menos a 2^160 osea (2^159) + 1
q = U | mask #se realiza una operacion OR entre U y la mascara
if is_prime(q, 20): #se verifica si q es primo de 20
break
# generate p
i = 0 # counter
j = 2 # offset
while i < 4096:
V = []
for k in range(n + 1):
arg = xmpz((s + j + k) % (2**g))
zzv = sha1(to_binary(arg)).hexdigest()
V.append(int(zzv, 16))
W = 0
for qq in range(0, n):
W += V[qq] * 2**(160 * qq)
W += (V[n] % 2**b) * 2**(160 * n)
X = W + 2**(L - 1)
c = X % (2 * q)
p = X - c + 1 # p = X - (c - 1)
if p >= 2**(L - 1):
if is_prime(p, 10):
return p, q
i += 1
j += n + 1
def generate_g(p, q):
while True:
h = randrange(2, p - 1)
exp = xmpz((p - 1) // q)
g = powmod(h, exp, p)
if g > 1:
break
return g
def generate_g(self,p,q): #生成g g=t^(p-1)/qmod p 1<t<p-1
while True:
h=randrange(2,p-1) #随机选取t 1<h<p-1 h^(p-1)/qmodp>1成立的整数
exp=xmpz((p-1)//q)
g=powmod(h,exp,p)
if g>1:
break
return g #生成的一个g=h^(p-1)/q
def sieve(limit=1000000):
sieve_limit = gmpy2.isqrt(limit) + 1
limit += 1
bitmap = gmpy2.xmpz(3)
bitmap[4:limit:2] = -1
for p in bitmap.iter_clear(3, sieve_limit):
bitmap[p * p:limit:p + p] = -1
return bitmap.iter_clear(2, limit)
def apply_bitmask_to_num(mask: str, num: int) -> int:
num = xmpz(num)
for i in range(36):
if mask[i] == '1':
num[i] = 1
elif mask[i] == '0':
num[i] = 0
return int(num)
def MonPro2 (X, Y): # Montgomery multiplication calc. (a_* b_) * r_ mod n
global n
global s
#global n_
#global r
n11 = modinv(n, 2) # r = 2
n11_ = 2 - n1
A = 0
m = s # (n.bit_length() or 1)
x = gmpy2.xmpz(X)
y = gmpy2.xmpz(Y)
for k in range(0, m):
a = gmpy2.xmpz(A)
q = mod((a[0] + x[k]*y[0])*n11_, 2) # pow(2, k)
A = (A + x[k]*Y + q*n) // 2
if (A >= n): # ??????
A = A - n
return A
def mkBitarray(self, index):
"""Produces a bitarray representing the presence / absence of the family in the pangenome using the provided index
The bitarray is stored in the :attr:`bitarray` attribute and is a :class:`gmpy2.xmpz` type.
:param index: The index computed by :func:`ppanggolin.pangenome.Pangenome.getIndex`
:type index: dict[:class:`ppanggolin.genome.Organism`, int]
"""
self.bitarray = gmpy2.xmpz(0) #pylint: disable=no-member
for org in self.organisms:
self.bitarray[index[org]] = 1
def ModExp (a, e, n): # Modular exponentiation with MonPro
global r
a_ = mod(a * r, n)
x_ = mod(1 * r, n)
j = (e.bit_length() or 1)
e = gmpy2.xmpz(e)
for i in range(j-1, -1, -1): # loop on i from j-1 downto 0
x_ = MonPro(x_, x_)
if (e[i] == 1):
x_ = MonPro(x_, a_)
return MonPro(x_, 1) # mod(x_ * r_, n)
def create_organisms_bits(dic_cluster_nb):
print("CREATE BITS...")
dic_organism_cluster = {}
for c in dic_cluster_nb:
for p in dic_cluster_nb[c]:
org = p.split("|")[0]
if org not in dic_organism_cluster:
dic_organism_cluster[org] = xmpz(0)
dic_organism_cluster[org][int(c)] = 1
return dic_organism_cluster
def update_tid_bitsets(model,data, attributes,tid_bitsets):
index_not_consider = xmpz(model.bitset_covered)
index_not_consider = list(index_not_consider.iter_set())
for i_at in attributes:
if attributes[i_at]["type"] == "numeric":
# delete cut points from tidbitset and attributes
for ncut in range(1,attributes[i_at]["ncutpoints"]+1):
del attributes[i_at][(i_at,ncut)]
del attributes[i_at][(i_at,-ncut)]
del tid_bitsets[(i_at,ncut)]
del tid_bitsets[(i_at,-ncut)]
init_bitset_numeric(data,attributes,i_at,tid_bitsets,*index_not_consider)
def generate_p_q(self):
g = self.N
n = (self.L - 1) // g
b = (L - 1) % g
while True:
while True:
s = xmpz(randrange(1, 2**(g)))
a = self.get_hash(s)
zz = xmpz((s + 1) % (2**g))
z = self.get_hash(zz)
U = int(a, 16) ^ int(z, 16)
mask = 2**(N - 1) + 1
q = U | mask
if is_prime(q, 20):
break
i, j = 0, 2
while i < 4096:
V = []
for k in range(n + 1):
arg = xmpz((s + j + k) % (2**g))
zzv = self.get_hash(arg)
V.append(int(zzv, 16))
W = 0
for qq in range(n):
W += V[qq] * 2**(160 * qq)
W += (V[n] % 2**b) * 2**(160 * n)
X = W + 2**(L - 1)
c = X % (2 * q)
p = X - c + 1
if p >= 2**(L - 1):
if is_prime(p, 10):
return p, q
i += 1
j += n + 1
def generate_p_q(L, N):
g = N # g >= 160
n = (L - 1) // g
b = (L - 1) % g
while True:
# generate q
while True:
s = xmpz(randrange(1, 2**(g)))
a = sha1(to_binary(s)).hexdigest()
zz = xmpz((s + 1) % (2**g))
z = sha1(to_binary(zz)).hexdigest()
U = int(a, 16) ^ int(z, 16)
mask = 2**(N - 1) + 1
q = U | mask
if is_prime(q, 20):
break
# generate p
i = 0 # counter
j = 2 # offset
while i < 4096:
V = []
for k in range(n + 1):
arg = xmpz((s + j + k) % (2**g))
zzv = sha1(to_binary(arg)).hexdigest()
V.append(int(zzv, 16))
W = 0
for qq in range(0, n):
W += V[qq] * 2**(160 * qq)
W += (V[n] % 2**b) * 2**(160 * n)
X = W + 2**(L - 1)
c = X % (2 * q)
p = X - c + 1 # p = X - (c - 1)
if p >= 2**(L - 1):
if is_prime(p, 10):
return p, q
i += 1
j += n + 1
def compute_statistic_numeric(model,tid):
statistic = dict()
# pattern related part
aux_bitset = xmpz(tid)
idx_bits = list(aux_bitset.iter_set())
values = model.target_values[idx_bits]
statistic["usage"] = values.size
if statistic["usage"] > 1:
#statistic["mean"],closest2,diff2 = compute_mean_and_twopoints(values,model.default_statistic["mean"])
statistic["mean"] = compute_mean(values)
closest2,diff2 = find2points(values,model.default_statistic["mean"])
statistic["mean2"] = compute_mean(closest2)
statistic["variance2"] = compute_RSS(closest2,statistic["mean2"])/2
statistic["RSS2"] = compute_RSS(closest2,model.default_statistic["mean"])
statistic["variance"] = compute_RSS(values,statistic["mean"])/statistic["usage"]
statistic["RSS_default_pattern"] = compute_RSS(values,model.default_statistic["mean"])
else:
statistic["mean"] = 0
statistic["variance"] = 0
statistic["variance2"] = 0
statistic["RSS2"] = 0
statistic["RSS_default_pattern"] = 0
# last rule related part
bitset_default = xmpz(model.bitset_uncovered &~ tid)
idx_bitsdef = list(bitset_default.iter_set())
values_uncovered = model.target_values[idx_bitsdef]
statistic["usage_default"] = len(values_uncovered)
if model.task == "discovery":
if statistic["usage_default"]:
statistic["RSS_default_uncovered"] =compute_RSS(values_uncovered,model.default_statistic["mean"])
else:
statistic["RSS_default_uncovered"] = 0
elif model.task == "classification":
statistic["mean_default"] = np.mean(values_uncovered)
statistic["variance_default"] = np.var(values_uncovered)
return statistic
def generate_p_q(self,L,N): #生成一个素数因子
g=N
n=(L-1)//g
b=(L-1)%g
while True:
while True:
s=xmpz(randrange(1,2**(g))) #生成一个大素数g
a=sha1(to_binary(s)).hexdigest()
zz=xmpz((s+1)%(2**g))
z = sha1(to_binary(zz)).hexdigest() #hash值为160bit，通过生成随机字符串将 | 链接起来新城新的字符串来实现p和q的生成
U = int(a, 16) ^ int(z, 16)
mask = 2 ** (N - 1) + 1
q = U | mask
if is_prime(q, 20): #is_prime用来判断是否为素数因子
break
# generate p #生成p
i = 0 # counter
j = 2 # offset
while i < 4096:
V = []
for k in range(n + 1):
arg = xmpz((s + j + k) % (2 ** g))
zzv = sha1(to_binary(arg)).hexdigest()
V.append(int(zzv, 16))
W = 0
for qq in range(0, n):
W += V[qq] * 2 ** (160 * qq)
W += (V[n] % 2 ** b) * 2 ** (160 * n)
X = W + 2 ** (L - 1)
c = X % (2 * q)
p = X - c + 1 # p = X - (c - 1)
if p >= 2 ** (L - 1):
if is_prime(p, 10):
return p, q #生成一个素数p:2^L-1<p<2^L,L为64的倍数 选取p-1的一个素数因子q,2^159<q<2^160
i += 1
j += n + 1
def generate_g(p, q):
while True:
h = randrange(
2, p - 1
) #esto es H en nuestro diagrama osea el numero aleatorio que deberia ser 2
exp = xmpz(
(p - 1) // q
) #esto es simplemente el cociente de 1024/160 = 6 que en nuestro diagrama es (P-1)/Q
#exp = ((p - 1) // q) #esto es simplemente el cociente de 1024/160 = 6 que en nuestro diagrama es (P-1)/Q
g = powmod(
h, exp,
p) #aqui se calcula G osea esto es simplemente (H^6) mod 1024
if g > 1:
break
return g
def decrypt(_c, _lambda, _m, _d, _mu, _n): c = gmpy2.xmpz(_c) lmda = gmpy2.xmpz(_lambda) m = gmpy2.xmpz(_m) d = gmpy2.xmpz(_d) mu = gmpy2.xmpz(_mu) n = gmpy2.xmpz(_n) b1 = f_mod(pow((f_mod(mul((((pow(c, lmda) \ % (pow(m, 2))-1))/m), mu), m)),d), n) return b1
def decrypt(_c, _lambda, _m, _d, _mu, _n): """ (M) = (((C^lambda mod (m^2)-1)/m)*mu mod m)^d mod n""" c = gmpy2.xmpz(_c) lmda = gmpy2.xmpz(_lambda) m = gmpy2.xmpz(_m) d = gmpy2.xmpz(_d) mu = gmpy2.xmpz(_mu) n = gmpy2.xmpz(_n) b1 = f_mod(pow((f_mod(mul((((pow(c, lmda) % (pow(m, 2))-1))/m), mu), m)),d), n) return b1
def encrypt(_g, _s, _e, _n, _m): r = gmpy2.xmpz(1) g = gmpy2.xmpz(_g) s = gmpy2.xmpz(_s) e = gmpy2.xmpz(_e) n = gmpy2.xmpz(_n) m = gmpy2.xmpz(_m) b1 = f_mod(e, n) b1 = pow(g, pow(s, b1)) b1 = mul(b1, f_mod(pow(r,m), pow(m,2))) return b1
def encrypt(_g, _s, _e, _n, _m): """C = (g^(M^(e mod n)))*((r^m)*mod (m^2))""" r = gmpy2.xmpz(1) g = gmpy2.xmpz(_g) s = gmpy2.xmpz(_s) e = gmpy2.xmpz(_e) n = gmpy2.xmpz(_n) m = gmpy2.xmpz(_m) b1 = f_mod(e, n) b1 = pow(g, pow(s, b1)) b1 = mul(b1, f_mod(pow(r,m), pow(m,2))) return b1
def computepi():
# N: number of decimals
f = StringIO()
w = xmpz(0)
k = 1
n1 = xmpz(4)
n2 = xmpz(3)
d = xmpz(1)
f10 = xmpz(10)
n10 = xmpz(-10)
i = 0
URL = request.args.get('URL', type=str)
N = request.args.get('N', default=10, type=int)
lenght = request.args.get('lenght', default=1, type=int)
while True:
# digit
u = int(div(n1, d))
v = int(div(n2, d))
if u == v:
f.write(chr(48 + u))
i += 1
if i % 10 == 0:
f.write("\t:%d\n" % i)
if i == N:
break
# extract
u = mul(d, mul(n10, u))
n1 = mul(n1, f10)
n1 = add(n1, u)
n2 = mul(n2, f10)
n2 = add(n2, u)
else:
# produce
k2 = k << 1
u = mul(n1, k2 - 1)
v = add(n2, n2)
w = mul(n1, k - 1)
n1 = add(u, v)
u = mul(n2, k + 2)
n2 = add(w, u)
d = mul(d, k2 + 1)
k += 1
if lenght != 0:
contents = urllib.request.urlopen(URL + "?N=" + str(N) + "&lenght=" +
str(lenght - 1)).read()
return (f.getvalue())
def apply_power_bitmask_to_num(mask: str, num: int) -> list:
queue = [xmpz(num)]
for i in range(36):
if mask[i] == '1':
for j in range(len(queue)):
queue[j][i] = 1
elif mask[i] == 'X':
new_queue = []
for ele in queue:
ele1 = ele.copy()
ele1[i] = 0
new_queue.append(ele1)
ele2 = ele.copy()
ele2[i] = 1
new_queue.append(ele2)
queue = new_queue
return [int(x) for x in queue]
def computepi():
# N: number of decimals
f = StringIO()
w = xmpz(0)
k = 1
n1 = xmpz(4)
n2 = xmpz(3)
d = xmpz(1)
f10 = xmpz(10)
n10 = xmpz(-10)
i = 0
N = request.args.get('N', default=10, type=int)
while True:
# digit
u = int(div(n1, d))
v = int(div(n2, d))
if u == v:
f.write(chr(48 + u))
i += 1
if i % 10 == 0:
f.write("\t:%d\n" % i)
if i == N:
break
# extract
u = mul(d, mul(n10, u))
n1 = mul(n1, f10)
n1 = add(n1, u)
n2 = mul(n2, f10)
n2 = add(n2, u)
else:
# produce
k2 = k << 1
u = mul(n1, k2 - 1)
v = add(n2, n2)
w = mul(n1, k - 1)
n1 = add(u, v)
u = mul(n2, k + 2)
n2 = add(w, u)
d = mul(d, k2 + 1)
k += 1
return (f.getvalue())
def add_rule(self, subgroup2add, tid_bitsets, attributes):
self.number_rules += 1
tid_cand = subgroup2add.bitset
self.bitset_covered = self.bitset_covered | tid_cand
self.support_covered = popcount(self.bitset_covered)
self.bitset_uncovered = self.bitset_uncovered & ~tid_cand
self.support_uncovered = popcount(self.bitset_uncovered)
self.bitset_rules.append(tid_cand)
self.antecedent_raw.append(subgroup2add.pattern)
self.statistic_rules.append(subgroup2add.statistic)
# IN CLASSIFICATION CASE EVERYTHING HAS TO BE UPDATED!
self.default_statistic["usage"] = self.support_uncovered
if self.task == "classification":
aux_tid = xmpz(self.bitset_uncovered)
idx_bitsdef = list(aux_tid.iter_set())
values_default = self.target_values[idx_bitsdef]
self.default_mean["mean"] = np.mean(values_default)
self.default_variance["variance"] = np.var(values_default)
# SHOULD BE REMOVED LATER ON
support = popcount(tid_cand)
self.support_rules.append(support)
self.add_pattern4prediction(subgroup2add.pattern, attributes)
self.consequent_description.append(
self.add_description_consequent(subgroup2add))
self.consequent_lastrule_description = self.add_consequent_lastrule()
self.add_description_antecedent(subgroup2add.pattern, attributes)
self.length_model = compute_length_model(self)
self.length_data = compute_length_data[self.target_type](self)
self.constant = delta_data_const[self.target_type](self)
if self.length_original > 0:
self.length_ratio = (self.length_data +
self.length_model) / self.length_original
elif self.length_original < 0:
self.length_ratio = self.length_original / (self.length_data +
self.length_model)
return self
def main():
# CSF - Use gmpy2's divmod instead of the Python built-in, it's slightly faster
divmod = f_divmod
bprint = sys.stdout.buffer.write
N = xmpz(int(sys.argv[1]))
# CSF - Used by bprint below to save a few usec off each print
line = "{:010d}\t:{}\n".format
# CSF - Not very PEP friendly, but the runtime on this benchmark is low, and
# this is faster than multiple single line assignments
n, a, d, t, u, i, k, ns, k1 = map(xmpz, (1, 0, 1, 0, 0, 0, 0, 0, 1))
while True:
k += 1
t = n << 1
n *= k
a += t
k1 += 2
a *= k1
d *= k1
if a >= n:
t, u = divmod(n * 3 + a, d)
u += n
if d > u:
ns = ns * 10 + t
i += 1
if not i % 10: # CSF - faster way of saying if i % 10 == 0
bprint(line(ns, i).encode())
ns = 0
if i >= N:
break
a -= d * t
a *= 10
n *= 10
def mkBitarray(self, index, partition='all'):
"""Produces a bitarray representing the presence / absence of the family in the pangenome using the provided index
The bitarray is stored in the :attr:`bitarray` attribute and is a :class:`gmpy2.xmpz` type.
:param index: The index computed by :func:`ppanggolin.pangenome.Pangenome.getIndex`
:type index: dict[:class:`ppanggolin.genome.Organism`, int]
:param partition: partition used to compute bitarray
:type partition: str
"""
self.bitarray = gmpy2.xmpz(0) # pylint: disable=no-member
if partition == 'all':
logging.getLogger().debug(f"all")
for org in self.organisms:
self.bitarray[index[org]] = 1
elif partition in ['shell', 'cloud']:
logging.getLogger().debug(f"shell, cloud")
if self.namedPartition == partition:
for org in self.organisms:
self.bitarray[index[org]] = 1
elif partition == 'accessory':
logging.getLogger().debug(f"accessory")
if self.namedPartition in ['shell', 'cloud']:
for org in self.organisms:
self.bitarray[index[org]] = 1
def sieve_gmpy2_iter(limit):
"""
Returns a list of the prime numbers up to limit (from 0 to limit).
A bit faster
https://gmpy2.readthedocs.io/en/latest/advmpz.html
"""
# Increment by 1 to account for the fact that slices do not include
# the last index value but we do want to include the last value for
# calculating a list of primes.
sieve_limit = gmpy2.isqrt(limit) + 1
limit += 1
# Mark bit positions 0 and 1 as not prime.
bitmap = gmpy2.xmpz(3)
# Process 2 separately. This allows us to use p+p for the step size
# when sieving the remaining primes.
bitmap[4:limit:2] = -1
# Sieve the remaining primes.
for p in bitmap.iter_clear(3, sieve_limit):
bitmap[p * p:limit:p + p] = -1
return bitmap.iter_clear(2, limit)
def sieve(limit=1000000):
'''credit to:
https://stackoverflow.com/questions/2897297/speed-up-bitstring-bit-operations-in-python
Returns a generator that yields the prime numbers up to limit.'''
# Increment by 1 to account for the fact that slices do not include
# the last index value but we do want to include the last value for
# calculating a list of primes.
sieve_limit = gmpy2.isqrt(limit) + 1
limit += 1
# Mark bit positions 0 and 1 as not prime.
bitmap = gmpy2.xmpz(3)
# Process 2 separately. This allows us to use p+p for the step size
# when sieving the remaining primes.
bitmap[4 : limit : 2] = -1
# Sieve the remaining primes.
for p in bitmap.iter_clear(3, sieve_limit):
bitmap[p*p : limit : p+p] = -1
return bitmap.iter_clear(2, limit)
def mkBitarray(self, index):
""" produces a bitarray representing the presence / absence of the family in the pangenome"""
self.bitarray = gmpy2.xmpz(0)
for org in self.organisms:
self.bitarray[index[org]] = 1
def plotTime(p):
n = (gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p))+(gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p)-1)
while not miller_rabin.millerRabin(n, 2):
n = (gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p))+(gmpy2.xmpz(RSA.getRandom())**gmpy2.xmpz(p)-1)
print ("number: %d and bit: %d"%(n, bit_length(n)))
for n in mersenne:
millis = int(round(time.time() * 1000000000000))
p = gmpy2.xmpz(n)
s = gmpy2.xmpz(2)
# (s**p)-1
millis = int(round(time.time() * 1000000000000)) - millis
print "time: %d"%millis
with open('file.txt', 'a') as f:
f.write("%d\n"%(n))
with open('mersenne.txt', 'a') as f:
f.write("%d\n"%(millis))
#countTime()
a = gmpy2.xmpz(1) # use 4 for good result
b = gmpy2.xmpz(500)
env.digitParameter = a
env.sample_string = " "
# Step 1
env._p = 29
env._q = 31
env._r = 37
env._s = 41
env.allNUmbers = []
# Step 2
env._n = 0
env._m = 0
env._phi = 0
env._lambda = 0
def generateLargePrime(p): n = (gmpy2.xmpz(getRandom())**gmpy2.xmpz(p))+(gmpy2.xmpz(getRandom())**gmpy2.xmpz(p)-1) while not miller_rabin.millerRabin(n, 2): n = (gmpy2.xmpz(getRandom())**gmpy2.xmpz(p))+(gmpy2.xmpz(getRandom())**gmpy2.xmpz(p)-1) return n
if algorithmRabin_Miller(n, 64) == 1: return 1 return 0 # print(algorithmRabin_Miller(887, 10)) if __name__ == '__main__': count_numbers = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0] i = 0 start = timeit.default_timer() start_total = start while True: if i == 100: break randbytes = os.urandom(384) randnumber = int.from_bytes(randbytes, 'big') randnumber = xmpz(randnumber) randnumber[0] = 1 randnumber[3071] = 1 # print(randnumber) if advanceRabin_Miller(randnumber) == 0: i += 1 stop = timeit.default_timer() took_time = stop - start start = timeit.default_timer() print(took_time) if 0 < took_time and took_time <= 1: count_numbers[0] += 1 elif 1 < took_time and took_time <= 2: count_numbers[1] += 1
# The Computer Language Benchmarks Game
# http://benchmarksgame.alioth.debian.org/
#
# contributed by Rene Bakker
# fixed by Isaac Gouy
import sys
from io import StringIO
from gmpy2 import xmpz, div, mul, add
N = int(sys.argv[1])
f = StringIO()
w = xmpz(0)
k = 1
n1 = xmpz(4)
n2 = xmpz(3)
d = xmpz(1)
f10 = xmpz(10)
n10 = xmpz(-10)
i = 0
while True:
# digit
u = int(div(n1, d))
v = int(div(n2, d))
if u == v:
f.write(chr(48 + u))
i += 1
if i % 10 == 0:
# end of sqr_n_mult
a1 = bit_width(2)
print(a1)
if __name__ == "__main__":
MM = 3, 7, 11, 29, 59, 107, 239
random.seed()
for M in range(0, 2**15):
if (M % 2 == 0):
continue
a = random.randint(0, M-1)
e = random.randint(0, M-1)
j = (e.bit_length() or 1)
e = gmpy2.xmpz(e)
x = sqr_n_mult2(a, e, M)
orig = pow(a, e, M)
if x != orig:
print("Error-B: n={0}; a={1}; e={2}; x={3}; orig={4};".format(M, a, e, x, orig))
# for n in range(pow(radix, start_bit), pow(radix, finish_bit)):
# if (n % 2 == 0):
# continue
# print("Output: n={0}".format(n))
#http://younglinux.info/python/task/even-odd
#https://habrahabr.ru/post/122538/
#http://forum.sources.ru/index.php?showtopic=348429
# partial unit test for gmpy2 xmpz functionality
# relies on Tim Peters' "doctest.py" test-driver
import gmpy2 as _g, doctest, sys, operator, gc
__test__={}
a=_g.xmpz(123)
b=_g.xmpz(456)
aa=_g.mpz(123)
bb=_g.mpz(456)
__test__['index']=\
r'''
>>> a=_g.xmpz(123)
>>> b=_g.xmpz(456)
>>> range(333)[a]
123
>>> range(333)[b]
Traceback (innermost last):
...
IndexError: range object index out of range
'''
__test__['elemop']=\
r'''
>>> a=_g.xmpz(123)
>>> b=_g.xmpz(456)
>>> a+b
mpz(579)
>>> a-b
mpz(-333)
>>> a*b