def make_train_csv_dict(train_csv):
train_dict = dict()
for author in train_csv:
train_dict[author['author_id']] = author
return train_dict
def make_clean_paper_author(confirmed_pairs, paper_author_pairs, output_file_path):
with open(output_file_path, 'w+') as f:
print >> f, 'PaperId,AuthorId,Name,Affiliation'
for idx in confirmed_pairs:
print >> f, '%d,%d,%s,%s' % (paper_author_pairs[idx]['paper_id'], paper_author_pairs[idx]['author_id'], paper_author_pairs[idx]['author_name'], paper_author_pairs[idx]['author_affiliation'])
if __name__ == '__main__':
if len(sys.argv) != 1 + 5:
print >> sys.stderr, "Usage : %s Train.csv PaperAuthor.csv Test.csv label output_file" % (sys.argv[0], )
sys.exit(-1)
train_csv = loader.load_train_csv(sys.argv[1])
train_csv_dict = make_train_csv_dict(train_csv)
paper_author_csv = loader.load_paper_author_csv(sys.argv[2])
test_csv = loader.load_test_csv(sys.argv[3])
labels = load_labels(sys.argv[4])
print 'Data loaded!'
confirmed_pairs = get_confirmed_pairs(get_author_in_train(train_csv),
train_csv_dict, test_csv, labels)
pair_idxs = filter_paper_author_csv(paper_author_csv, confirmed_pairs)
make_clean_paper_author(pair_idxs, paper_author_csv, sys.argv[5])
sys.exit(0)
from python_util.config import *
def get_internal_author_ids(internal_train_csv):
author_ids = dict()
for rec in internal_train_csv:
author_ids[rec['author_id']] = 1
return author_ids
if __name__ == '__main__':
if len(sys.argv) != 1 + 3:
print >> sys.stderr, "Usage : %s feature_file internal_train_output internal_valid_output" % (sys.argv[0],)
sys.exit(-1)
train_csv = loader.load_train_csv(DATA['Train'])
internal_train_csv = loader.load_train_csv(DATA['internal_train'])
internal_valid_csv = loader.load_test_csv(DATA['internal_valid'])
author_papers = dict()
with open(sys.argv[1], 'r') as f:
lines = f.read().split('\n')
pair_count = 0
for rec in train_csv:
author_id = rec['author_id']
paper_ids = rec['confirmed_paper_ids'] + rec['deleted_paper_ids']
paper_id_count = len(paper_ids)
author_papers[author_id] = dict()
def get_internal_author_ids(internal_train_csv):
author_ids = dict()
for rec in internal_train_csv:
author_ids[rec['author_id']] = 1
return author_ids
if __name__ == '__main__':
if len(sys.argv) != 1 + 3:
print >> sys.stderr, "Usage : %s feature_file internal_train_output internal_valid_output" % (
sys.argv[0], )
sys.exit(-1)
train_csv = loader.load_train_csv(DATA['Train'])
internal_train_csv = loader.load_train_csv(DATA['internal_train'])
internal_valid_csv = loader.load_test_csv(DATA['internal_valid'])
author_papers = dict()
with open(sys.argv[1], 'r') as f:
lines = f.read().split('\n')
pair_count = 0
for rec in train_csv:
author_id = rec['author_id']
paper_ids = rec['confirmed_paper_ids'] + rec['deleted_paper_ids']
paper_id_count = len(paper_ids)
author_papers[author_id] = dict()
for idx, paper_id in enumerate(paper_ids):
def make_clean_paper_author(confirmed_pairs, paper_author_pairs,
output_file_path):
with open(output_file_path, 'w+') as f:
print >> f, 'PaperId,AuthorId,Name,Affiliation'
for idx in confirmed_pairs:
print >> f, '%d,%d,%s,%s' % (
paper_author_pairs[idx]['paper_id'],
paper_author_pairs[idx]['author_id'],
paper_author_pairs[idx]['author_name'],
paper_author_pairs[idx]['author_affiliation'])
if __name__ == '__main__':
if len(sys.argv) != 1 + 5:
print >> sys.stderr, "Usage : %s Train.csv PaperAuthor.csv Test.csv label output_file" % (
sys.argv[0], )
sys.exit(-1)
train_csv = loader.load_train_csv(sys.argv[1])
train_csv_dict = make_train_csv_dict(train_csv)
paper_author_csv = loader.load_paper_author_csv(sys.argv[2])
test_csv = loader.load_test_csv(sys.argv[3])
labels = load_labels(sys.argv[4])
print 'Data loaded!'
confirmed_pairs = get_confirmed_pairs(get_author_in_train(train_csv),
train_csv_dict, test_csv, labels)
pair_idxs = filter_paper_author_csv(paper_author_csv, confirmed_pairs)
make_clean_paper_author(pair_idxs, paper_author_csv, sys.argv[5])
sys.exit(0)