def create_hashed_features_wdk(param, data):
"""
creates hashed dot features for the wdk
"""
# fix parameters
start_degree = 0
hash_bits = 12
degree = param["degree"]
order = 1
gap = 0
reverse = True
#print "test", data[0]
# create raw features
feats_char = StringCharFeatures(data, DNA)
feats_raw = StringByteFeatures(DNA)
feats_raw.obtain_from_char(feats_char, order - 1, order, gap, reverse)
# finish up
feats = HashedWDFeaturesTransposed(feats_raw, start_degree, degree, degree,
hash_bits)
#feats = HashedWDFeatures(feats_raw, start_degree, degree, degree, hash_bits)
#feats = WDFeatures(feats_raw, 1, 8)#, degree, hash_bits)
return feats
def features_snp_modular(fname): from shogun.Features import StringByteFeatures, SNPFeatures, SNP sf=StringByteFeatures(SNP) sf.load_ascii_file(fname, False, SNP, SNP) #print sf.get_features() snps=SNPFeatures(sf)
def features_snp_modular(fname):
from shogun.Features import StringByteFeatures, SNPFeatures, SNP
sf = StringByteFeatures(SNP)
sf.load_ascii_file(fname, False, SNP, SNP)
#print sf.get_features()
snps = SNPFeatures(sf)
def create_hashed_features_wdk(data, degree):
"""
creates hashed dot features for the wdk
"""
# fix parameters
start_degree = 0
hash_bits = 12
order = 1
gap = 0
reverse = True
dat = [str(xt) for xt in data]
# create raw features
feats_char = StringCharFeatures(dat, DNA)
feats_raw = StringByteFeatures(DNA)
feats_raw.obtain_from_char(feats_char, order - 1, order, gap, reverse)
# finish up
feats = HashedWDFeaturesTransposed(feats_raw, start_degree, degree, degree,
hash_bits)
#feats = HashedWDFeatures(feats_raw, start_degree, degree, degree, hash_bits)
#feats = WDFeatures(feats_raw, 1, 8)#, degree, hash_bits)
return feats
def create_hashed_features_wdk(param, data):
"""
creates hashed dot features for the wdk
"""
# fix parameters
start_degree = 0
hash_bits = 12
degree = param["degree"]
order = 1
gap = 0
reverse = True
#print "test", data[0]
# create raw features
feats_char = StringCharFeatures(data, DNA)
feats_raw = StringByteFeatures(DNA)
feats_raw.obtain_from_char(feats_char, order-1, order, gap, reverse)
# finish up
feats = HashedWDFeaturesTransposed(feats_raw, start_degree, degree, degree, hash_bits)
#feats = HashedWDFeatures(feats_raw, start_degree, degree, degree, hash_bits)
#feats = WDFeatures(feats_raw, 1, 8)#, degree, hash_bits)
return feats
def distribution_ppwm_modular (fm_dna=traindna, order=3):
from shogun.Features import StringByteFeatures, StringCharFeatures, DNA
from shogun.Distribution import PositionalPWM
from numpy import array,e,log,exp
charfeat=StringCharFeatures(DNA)
charfeat.set_features(fm_dna)
feats=StringByteFeatures(charfeat.get_alphabet())
feats.obtain_from_char(charfeat, order-1, order, 0, False)
L=20
k=3
sigma = 1;
mu = 4
ppwm=PositionalPWM()
ppwm.set_sigma(sigma)
ppwm.set_mean(mu)
pwm=array([[0.0, 0.5, 0.1, 1.0],
[0.0, 0.5, 0.5, 0.0],
[1.0, 0.0, 0.4, 0.0],
[0.0, 0.0, 0.0, 0.0]]);
pwm=array([[0.01,0.09,0.1],[0.09,0.01,0.1],[0.85,0.4,0.1],[0.05,0.5,0.7]])
ppwm.set_pwm(log(pwm))
#print(ppwm.get_pwm())
ppwm.compute_w(L)
w=ppwm.get_w()
#print(w)
#from pylab import *
#figure(1)
#pcolor(exp(w))
#pcolor(w)
#colorbar()
#figure(2)
ppwm.compute_scoring(1)
u=ppwm.get_scoring(0)
#pcolor(exp(u))
#show()
#ppwm=PositionalPWM(feats)
#ppwm.train()
#out_likelihood = histo.get_log_likelihood()
#out_sample = histo.get_log_likelihood_sample()
return w,u
def features_string_hashed_wd_modular(A=matrix,order=3,start_order=1,hash_bits=2):
a=LongIntFeatures(A)
from numpy import array, uint8
from shogun.Features import HashedWDFeatures, StringByteFeatures, RAWDNA
from shogun.IO import MSG_DEBUG
x=[array([0,1,2,3,0,1,2,3,3,2,2,1,1],dtype=uint8)]
from_order=order
f=StringByteFeatures(RAWDNA)
#f.io.set_loglevel(MSG_DEBUG)
f.set_features(x)
y=HashedWDFeatures(f,start_order,order,from_order,hash_bits)
fm=y.get_computed_dot_feature_matrix()
return fm
def distribution_ppwm_modular(fm_dna=traindna, order=3):
from shogun.Features import StringByteFeatures, StringCharFeatures, DNA
from shogun.Distribution import PositionalPWM
from numpy import array, e, log, exp
charfeat = StringCharFeatures(DNA)
charfeat.set_features(fm_dna)
feats = StringByteFeatures(charfeat.get_alphabet())
feats.obtain_from_char(charfeat, order - 1, order, 0, False)
ppwm = PositionalPWM()
ppwm.set_sigma(5.0)
ppwm.set_mean(10.0)
pwm = array([[0.0, 0.5, 0.1, 1.0], [0.0, 0.5, 0.5, 0.0], [1.0, 0.0, 0.4, 0.0], [0.0, 0.0, 0.0, 0.0]])
ppwm.set_pwm(log(pwm))
print ppwm.get_pwm()
ppwm.compute_w(20)
w = ppwm.get_w()
def distribution_ppwm_modular (fm_dna=traindna, order=3):
from shogun.Features import StringByteFeatures, StringCharFeatures, DNA
from shogun.Distribution import PositionalPWM
from numpy import array,e,log,exp
charfeat=StringCharFeatures(DNA)
charfeat.set_features(fm_dna)
feats=StringByteFeatures(charfeat.get_alphabet())
feats.obtain_from_char(charfeat, order-1, order, 0, False)
ppwm=PositionalPWM()
ppwm.set_sigma(5.0)
ppwm.set_mean(10.0)
pwm=array([[0.0, 0.5, 0.1, 1.0],
[0.0, 0.5, 0.5, 0.0],
[1.0, 0.0, 0.4, 0.0],
[0.0, 0.0, 0.0, 0.0]]);
ppwm.set_pwm(log(pwm))
print ppwm.get_pwm()
ppwm.compute_w(20)
w= ppwm.get_w()