def _fingerprint_similarity( self, f1, f2 ):
print "computing similarity for fingerprints: {} {}".format(f1, f2)
f1,f2= DictTree.intersect( f1, f2 )
def feature_map(*features):
f1,f2= features
return f1.similarity(f2)
similarities= DictTree.map( feature_map, f1, f2 )
return self._reducer( similarities )
def _fingerprint_similarity(self, f1, f2):
print "computing similarity for fingerprints: {} {}".format(
f1, f2)
f1, f2 = DictTree.intersect(f1, f2)
def feature_map(*features):
f1, f2 = features
return f1.similarity(f2)
similarities = DictTree.map(feature_map, f1, f2)
return self._reducer(similarities)
def visualize_normal_composites( composites, show=True ):
COLORS= ('red', 'orange','yellow','green', 'blue', 'violet')
n= len(composites)
width= 0.9-(0.1*n)
common= DictTree.intersect( *composites )
colors= COLORS[:n]
for i,c,color in zip(range(n),common, colors):
visualize_normal_composite( c, color=color, show=False, offset=i*0.1, width=width)
if show:
pyplot.show()
def fit( self, data, labels=None ):
def feature_map(*features):
assert len(features)==1
f= features[0]
try:
if isinstance( f, FloatSeq ):
return GaussianAnomalyModel.from_features( f.name, f.data )
else:
raise Exception("Unknown feature: {}".format(f))
except InsufficientData:
return self.IGNORE_CHILD
assert isinstance( data, CompositeFeature)
newmodel= DictTree.map( feature_map, data)
self.clear()
self.update( newmodel )
def fit(self, data, labels=None):
def feature_map(*features):
assert len(features) == 1
f = features[0]
try:
if isinstance(f, FloatSeq):
return GaussianAnomalyModel.from_features(f.name, f.data)
else:
raise Exception("Unknown feature: {}".format(f))
except InsufficientData:
return self.IGNORE_CHILD
assert isinstance(data, CompositeFeature)
newmodel = DictTree.map(feature_map, data)
self.clear()
self.update(newmodel)