def train_classifier():
print "Training the classifier"
system_classifier = Classifier.objects.first()
system_classifier.status = "training"
system_classifier.save()
print "Fetching Chunks"
chunks = Chunk.get_chunks()
authors = []
fingerprints = []
for chunk in chunks:
if chunk.author:
authors.append(chunk.author.name)
fingerprints.append(chunk.get_fingerprint_list())
print "Training..."
clf = classifier.svm.train_svm(fingerprints, authors)
if clf:
print "Trained the classifier"
store_trained_classifier.delay(clf)
return True
else:
print "Failed to train classifier"
system_classifier.status = "untrained"
system_classifier.save()
return False
def create_text_average_chunk(text_id):
text = Text.objects.get(pk=text_id)
if (text.average_chunk is not None):
text.average_chunk.delete()
chunks = Chunk.get_chunks().filter(text=text)
print "averaging %i chunks" % (len(chunks))
average_fingerprint = Chunk.get_average_fingerprint_of_chunks(chunks)
chunk = Chunk.objects.create()
for key in average_fingerprint.keys():
setattr(chunk, key, average_fingerprint[key])
chunk.text = text
chunk.save()
text.average_chunk = chunk
text.save()
return True
