def run(self):
seed_filename = self.pass_in
output_filename = self.pass_out
status_count_threshold = self.param['status_count_threshold']
following_count_threshold = self.param['following_count_threshold']
seeds = dp_seeds.fromCSVThresholdedEng(seed_filename,
status_count_threshold,
following_count_threshold)
fout = open(output_filename, 'w')
for user_id, age_group in seeds:
fout.write(str(user_id) + '\t' + str(age_group) + '\n')
fout.close()
from dpFeature import Feature
import dpAnnotatedSeeds as dp_seeds
from dpDataLoader import loadFollowingFromDBofUser
class FollowingFeature(Feature):
def convertSingle(self, user_id):
friends = loadFollowingFromDBofUser(user_id)
def _dummy(friend): # insert feature weight
return (friend, 1)
friends = map(_dummy, friends)
return friends
if __name__ == "__main__":
seed_file = '../data/sampled2.8-annotated.csv'
seeds = dp_seeds.fromCSVThresholdedEng(seed_file)
feature_gen = FollowingFeature(seeds)
fout_name = '../feature_set2/ver2.8-efollowing.libsvm'
feature_gen.generateTo(fout_name)
