def __init__(self,name):
self.metric=name
self.make_name()
#self.param=0
self.simCalc=SimCalculator()
#self.reverse=False
self.param=(0,False) #flag for negating values to swap direction of inequality - false is >, true is <
class ClassifierUP():
def __init__(self,name):
self.metric=name
self.make_name()
#self.param=0
self.simCalc=SimCalculator()
#self.reverse=False
self.param=(0,False) #flag for negating values to swap direction of inequality - false is >, true is <
def make_name(self):
self.name=self.metric+"_UP"
def fit(self,pairs,term_map,target):
logging.info("Baseline fit: Ignoring training data as unsupervised classifier: "+self.name)
def predict(self,pairs,term_map):
#term_map is dictionary from terms (in pairs) to vectors
#print "Baseline prediction: "+self.name
#print "Generating width_map from "+str(len(term_map.keys()))+" keys"
tags=[]
for pair in pairs:
wd = self.simCalc.compute_score(pair,term_map,self.metric)
if self.param[1]:
wd=-wd
if wd > self.param[0]:
tags.append(1)
else:
tags.append(0)
return np.array(tags,dtype=int)
def get_params(self):
return str(self.param)
class ClassifierUP():
def __init__(self,name):
self.metric=name
self.make_name()
self.param=0
self.simCalc=SimCalculator()
def make_name(self):
self.name=self.metric+"_UP"
def fit(self,pairs,term_map,target):
logging.info("Baseline fit: Ignoring training data as unsupervised classifier: "+self.name)
def predict(self,pairs,term_map):
#term_map is dictionary from terms (in pairs) to vectors
#print "Baseline prediction: "+self.name
#print "Generating width_map from "+str(len(term_map.keys()))+" keys"
tags=[]
for pair in pairs:
wd = self.simCalc.compute_score(pair,term_map,self.metric)
if wd > self.param:
tags.append(1)
else:
tags.append(0)
return np.array(tags,dtype=int)
def __init__(self,name):
self.metric=name
self.make_name()
self.param=0
self.simCalc=SimCalculator()
