def extract_raw_refs_from_pdfs():
metas = util.load_json(cfg.paths['papers-metadata'])
refs = util.parallelize(extract_raw_refs_from_pdf, metas, N_THREADS)
refs = {m['uid']: r for m, r in zip(metas, refs)}
util.save_json(cfg.paths['raw-papers-refs'], refs)
print('saved raw papers refs to "{}"'.format(cfg.paths['raw-papers-refs']))
def process():
cands = load_results()
model = AttentionModel()
mc_ctx = MultiChoiceQuestionManger(subset='val')
results = []
t = time()
for i, res_i in enumerate(cands):
if i % 100 == 0:
avg_time = (time() - t) / 100.
print('%d/%d (%0.2f sec/sample)' % (i, len(cands), avg_time))
t = time()
image_id = res_i['image_id']
aug_id = res_i['aug_id']
question = res_i['target']
# question_id = int(aug_id / 1000)
question_id = res_i['question_id']
gt_answer = mc_ctx.get_gt_answer(question_id)
pred_ans, sc = model.get_score(image_id, question)
is_valid = compare_answer(pred_ans, gt_answer)
# import pdb
# pdb.set_trace()
if not is_valid:
continue
t_i = {
'image_id': int(image_id),
'aug_id': aug_id,
'question_id': question_id,
'question': question,
'score': float(sc)
}
results.append(t_i)
save_json('result/bs_vis_scores_mlb2-att.json', results)
def get_journals():
pos, cnt = 1, 0
util.mkdir(JOURNAL_FOLDER)
while True:
html = util.get_page(JOURNAL_URL + str(pos))
links = util.find_journals(html)
once_cnt = 0
for link in links:
if link[0] == '' or '?' in link[0]:
continue
data = {}
data['type'] = 'journal'
data['short'] = link[0]
data['name'] = link[1]
data['url'] = 'http://dblp.uni-trier.de/db/journals/' + data[
'short']
util.save_json(
os.path.join(JOURNAL_FOLDER, util.hex_hash(data['short'])),
data)
cnt += 1
once_cnt += 1
if once_cnt == 0:
break
pos += 100
print 'Journal', cnt
def get_conferences():
files = util.listdir(CONFERENCE_FOLDER)
util.mkdir(CONFERENCE_CRALWED_FOLDER)
cnt = 0
conf = util.load_json('conf_name.json')
for file_name in files:
save_path = os.path.join(CONFERENCE_CRALWED_FOLDER, file_name)
if util.exists(save_path):
continue
data = util.load_json(os.path.join(CONFERENCE_FOLDER, file_name))
if data['short'] not in conf.keys():
continue
html = util.get_page(data['url'])
subs = get_subs(data['short'], html)
data['name'] = conf[data['short']]
data['sub'] = {}
for sub in subs:
if sub not in conf.keys():
continue
html = util.get_page('http://dblp.uni-trier.de/db/conf/' + sub)
data['sub'][sub] = {}
data['sub'][sub]['pub'] = get_publications(html)
data['sub'][sub]['name'] = conf[sub]
cnt += 1
print cnt, len(files), data['short']
util.save_json(save_path, data)
def process(model_type='mlb'):
cands = load_results()
if model_type == 'mlb':
model = AttentionModel()
else:
model = VanillaModel()
mc_ctx = MultiChoiceQuestionManger(subset='val')
results = {}
t = time()
for i, res_key in enumerate(cands):
if i % 100 == 0:
avg_time = (time() - t) / 100.
print('%d/%d (%0.2f sec/sample)' % (i, len(cands), avg_time))
t = time()
res_i = cands[res_key]
image_id = res_i['image_id']
question = res_i['target']
question_id = res_i['question_id']
gt_answer = mc_ctx.get_gt_answer(question_id)
pred_ans, scores = model.get_score(image_id, question)
sc = float(scores)
is_valid = compare_answer(pred_ans, gt_answer)
# if not is_valid:
# continue
results[res_key] = {
'pred_answer': pred_ans,
'pred_score': sc,
'gt_answer': gt_answer,
'is_valid': is_valid
}
save_json('result/%s_scores_final_v2.json' % model_type, results)
def save(self, folder_name="autosave"):
base_path = os.path.join(self.save_path, folder_name)
print 'save %s to %s' % (self.scene_handler.scene.name, base_path)
scn = self.scene_handler.scene
util.mkdir_if_absent(base_path)
util.save_json(self.dict_repr(), os.path.join(base_path, 'game'))
util.save_json(scn.dict_repr(), os.path.join(base_path, scn.name))
def process():
def _parse_image_id(image):
return int(image.split('.')[0].split('_')[-1])
model = AttentionModel()
ans2top_ans = AnswerTokenToTopAnswer()
task_data_dir = '/usr/data/fl302/code/utils/bs_data_maker'
task_data_file = os.path.join(task_data_dir, 'task_data_for_verif.json')
task_data = load_json(task_data_file)
is_valid = []
num = len(task_data)
for i, info in enumerate(task_data):
print('%d/%d' % (i, num))
image = info['image']
image_id = _parse_image_id(image)
question = info['target']
answer = info['answer']
scores = model.inference(image_id, question)
scores[:, -1] = -10.
# pdb.set_trace()
top_ans_id = ans2top_ans.direct_query(answer)
if top_ans_id == 2000:
raise Exception('Warning: answer oov')
scores = scores.flatten()
pred_top_ans_id = scores.argmax()
is_valid.append(int(pred_top_ans_id == top_ans_id))
n_valid = sum(is_valid)
print('valid: %d/%d' % (n_valid, num))
save_json(os.path.join(task_data_dir, 'task_data_verif_state.json'),
is_valid)
def get_conferences():
files = util.listdir(CONFERENCE_FOLDER)
util.mkdir(CONFERENCE_CRALWED_FOLDER)
cnt = 0
conf = util.load_json('conf_name.json')
for file_name in files:
save_path = os.path.join(CONFERENCE_CRALWED_FOLDER, file_name)
if util.exists(save_path):
continue
data = util.load_json(os.path.join(CONFERENCE_FOLDER, file_name))
if data['short'] not in conf.keys():
continue
html = util.get_page(data['url'])
subs = get_subs(data['short'], html)
data['name'] = conf[data['short']]
data['sub'] = {}
for sub in subs:
if sub not in conf.keys():
continue
html = util.get_page('http://dblp.uni-trier.de/db/conf/' + sub)
data['sub'][sub] = {}
data['sub'][sub]['pub'] = get_publications(html)
data['sub'][sub]['name'] = conf[sub]
cnt += 1
print cnt, len(files), data['short']
util.save_json(save_path, data)
def score_replay_buffer():
d = load_json('vqa_replay_buffer/low/vqa_replay.json')
vqa_data = VQAData()
# create model
sess, model = create_model()
memory = d['memory']
new_memory = {}
for i, quest_key in enumerate(memory.keys()):
pathes = memory[quest_key]
if len(pathes) == 0:
continue
if i % 100 == 0:
print('Processed %d/%d items' % (i, len(memory)))
new_memory[quest_key] = {}
# if it has valid questions
quest_id = int(quest_key)
pathes = memory[quest_key].keys()
quest, quest_len = process_questions(pathes)
image, top_ans = vqa_data.get_data(quest_id)
vqa_inputs = [image, quest, quest_len, top_ans]
scores = sess.run(model.prob,
feed_dict=model.fill_feed_dict(vqa_inputs))
confs = scores[:, top_ans]
ref = memory[quest_key]
for path_key, new_conf in zip(pathes, confs):
new_memory[quest_key][path_key] = (ref[path_key], float(new_conf))
# save
save_json('vqa_replay_buffer/vqa_replay_low_rescore.json',
{'memory': new_memory})
def process_worker(subset, id, proc_range, to_sentence):
# params
k = 50
res_file = 'result/quest_vaq_nn_%s_worker%d.json' % (subset, id)
# load distances
val_qids, nn_ids = load_image_nn(subset=subset)
# slice
val_qids = val_qids[proc_range]
nn_ids = nn_ids[proc_range]
# create nn model
nn_model = QuestionPool()
# run
num = len(val_qids)
results = []
for i, (v_qid, v_nn) in enumerate(zip(val_qids, nn_ids)):
# run nn search
t = time()
tr_qid, tr_path = nn_model.get_candidates(v_nn[:k])
sent = to_sentence.index_to_question(tr_path)
print(sent)
print('P%d: Processing %d/%d, time %0.2f sec.' %
(id, i, num, time() - t))
res_i = {'question_id': int(v_qid), 'question': sent}
results.append(res_i)
save_json(res_file, results)
print('P%d: Done' % id)
return
def process():
cands = load_results()
model = N2MNWrapper()
mc_ctx = MultiChoiceQuestionManger(subset='val')
results = []
t = time()
for i, res_i in enumerate(cands):
if i % 100 == 0:
avg_time = (time() - t) / 100.
print('%d/%d (%0.2f sec/sample)' % (i, len(cands), avg_time))
t = time()
image_id = res_i['image_id']
aug_id = res_i['question_id']
question = res_i['question']
question_id = int(aug_id / 1000)
gt_answer = mc_ctx.get_gt_answer(question_id)
pred_answers, scores = model.inference(image_id, [question])
sc = scores[0]
pred_ans = pred_answers[0]
is_valid = compare_answer(pred_ans, gt_answer)
# import pdb
# pdb.set_trace()
if not is_valid:
continue
t_i = {
'image_id': int(image_id),
'question_id': aug_id,
'question': question,
'score': float(sc)
}
results.append(t_i)
save_json('result/vae_ia_van_n2mn_flt_full.json', results)
def save(self, folder_name="autosave"):
base_path = os.path.join(self.save_path, folder_name)
print 'save %s to %s' % (self.scene_handler.scene.name, base_path)
scn = self.scene_handler.scene
util.mkdir_if_absent(base_path)
util.save_json(self.dict_repr(), os.path.join(base_path, 'game'))
util.save_json(scn.dict_repr(), os.path.join(base_path, scn.name))
def precompute(num_p, offset=0, primes_path="primes.txt", store_dir=store_dir):
Path(store_dir).mkdir(exist_ok=True)
primes = load_primes(primes_path)
for p in tqdm(primes[offset:num_p], desc="Precomputing"):
rc_dict = precompute_p(p)
util.save_json(p, rc_dict, store_dir)
def merge_result(res1, res2):
results = []
unk_counts = []
batch_vqa_scores = []
for quest_id in res1:
items1 = res1[quest_id]
items2 = res2[quest_id]
res_i, scores_i, n_i = _merge_item(items1, items2)
results += res_i
batch_vqa_scores.append(scores_i.mean())
unk_counts.append(n_i)
# save results
res_file = 'result/bs_gen_%s.json' % 'vae_ia_merge'
score_file = 'result/bs_vqa_scores_%s.mat' % 'vae_ia_merge'
save_json(res_file, results)
batch_vqa_scores = np.array(batch_vqa_scores, dtype=np.float32)
mean_vqa_score = batch_vqa_scores.mean()
mean_unk_count = np.mean(unk_counts)
savemat(score_file, {
'scores': batch_vqa_scores,
'mean_score': mean_vqa_score
})
print('BS mean VQA score: %0.3f' % mean_vqa_score)
print('BS mean #questions: %0.3f' % mean_unk_count)
return res_file, mean_vqa_score, mean_unk_count
def process():
# load data
blacklist = make_blacklist()
save_json('data/kptest_blacklist.json', blacklist)
qrpe = load_qrpe_data(blacklist)
vtfp = load_vtfp_data(blacklist)
import pdb
pdb.set_trace()
meta = qrpe + vtfp
# process data
images, image_ids, questions = [], [], []
encoder = SentenceEncoder()
for item in meta:
image_id = item['image_id']
image = item['image']
tokens = encoder.encode_sentence(item['question'])
images.append(image)
image_ids.append(image_id)
questions.append(tokens)
# put to array
from post_process_variation_questions import put_to_array
arr, arr_len = put_to_array(questions)
save_json('data/QRD_irrelevant_meta.json', {
'images': images,
'image_ids': image_ids
})
image_ids = np.array(image_ids, dtype=np.int32)
save_hdf5('data/QRD_irrelevant_data.data', {
'image_ids': image_ids,
'quest': arr,
'quest_len': arr_len
})
def process_test():
from util import save_hdf5, save_json
# load data
meta = load_bsir_dataset()
# process data
labels, images, image_ids, questions = [], [], [], []
encoder = SentenceEncoder()
for item in meta:
image_id = item['image_id']
image = item['image']
tokens = encoder.encode_sentence(item['question'])
images.append(image)
image_ids.append(image_id)
questions.append(tokens)
labels.append(item['label'])
# put to array
from post_process_variation_questions import put_to_array
arr, arr_len = put_to_array(questions)
save_json('data/QRD_irrelevant_meta_test.json', {
'images': images,
'image_ids': image_ids
})
image_ids = np.array(image_ids, dtype=np.int32)
labels = np.array(labels, dtype=np.float32)
import pdb
pdb.set_trace()
save_hdf5(
'data/QRD_irrelevant_data_test.data', {
'image_ids': image_ids,
'quest': arr,
'quest_len': arr_len,
'labels': labels
})
def _encode_w2v(images, encoder, subset):
quest_coding = []
cands_coding = []
labels = []
quest_ids = []
cands_meta = []
for i, info in enumerate(images):
if not i % 1000:
tf.logging.info("%s: processed %d of %d items." %
(subset.upper(), i, len(images)))
quest_id = info.question_id
q_w2v = encoder.encode(info.question)
ca_w2v, label = _encode_answer_candidates(info, encoder)
# pdb.set_trace()
quest_coding.append(q_w2v)
cands_coding.append(ca_w2v)
labels.append(label)
quest_ids.append(quest_id)
_m = {'quest_id': quest_id, 'cands': info.choices}
cands_meta.append(_m)
# ready to pack data
quest_coding = np.concatenate(quest_coding, axis=0).astype(np.float32)
cands_coding = np.concatenate(cands_coding, axis=0).astype(np.float32)
labels = np.array(labels, dtype=np.int32)
quest_ids = np.array(quest_ids, dtype=np.int32)
save_hdf5(
'data3/vqa_mc_w2v_coding_%s.data' % subset, {
'quest_w2v': quest_coding,
'cands_w2v': cands_coding,
'labels': labels,
'quest_ids': quest_ids
})
save_json('data3/vqa_mc_cands_%s.meta' % subset, cands_meta)
def search_pub_and_save(uid, query, base_dir):
data = {
'uid': uid,
'query': query,
'success': True,
'message': '',
'result': {},
}
path = get_search_result_path(uid, query, base_dir)
if os.path.isfile(path):
data = util.load_json(path)
if data['success'] and data['result']:
print('skipping query "{}" with results'.format(query))
return
print('executing query "{}"'.format(query))
try:
result = search_pub(query)
data['result'] = result
except Exception as e:
error_msg = str(e)
print('ERROR for query "{}": "{}"'.format(query, error_msg))
data['success'] = False
data['message'] = error_msg
util.save_json(path, data)
print('saved query "{}" to {}'.format(query, path))
global sleep_time
if not data['result']:
sleep_time *= 1.1
print('sleeping for {:.2f} seconds'.format(sleep_time))
time.sleep(sleep_time)
return data
def parse_raw_refs():
refs = util.load_json(cfg.paths['raw-papers-refs'])
data = util.parallelize(_parse_raw_refs, list(refs.values()), N_THREADS)
refs = {k: v for k, v in zip(refs.keys(), data)}
util.save_json(cfg.paths['papers-refs'], refs)
print('saved papers refs to "{}"'.format(cfg.paths['papers-refs']))
def process():
cands = load_results()
model = N2MNWrapper()
mc_ctx = MultiChoiceQuestionManger(subset='val')
results = {}
t = time()
for i, res_key in enumerate(cands):
if i % 100 == 0:
avg_time = (time() - t) / 100.
print('%d/%d (%0.2f sec/sample)' % (i, len(cands), avg_time))
t = time()
res_i = cands[res_key]
image_id = res_i['image_id']
question = res_i['target']
question_id = res_i['question_id']
gt_answer = mc_ctx.get_gt_answer(question_id)
pred_answers, scores = model.inference(image_id, [question])
sc = float(scores[0])
pred_ans = pred_answers[0]
is_valid = compare_answer(pred_ans, gt_answer)
# if not is_valid:
# continue
results[res_key] = {
'pred_answer': pred_ans,
'pred_score': sc,
'gt_answer': gt_answer,
'is_valid': is_valid
}
save_json('result/n2mn_scores_final_v2.json', results)
def ivqa_decoding_beam_search(checkpoint_path=None, subset='kpval'):
model_config = ModelConfig()
res_file = 'result/quest_vaq_greedy_%s_%s.json' % (
FLAGS.model_type.upper(), subset)
# Get model
model_fn = get_model_creation_fn(FLAGS.model_type)
create_fn = create_reader(FLAGS.model_type, phase='test')
# Create the vocabulary.
to_sentence = SentenceGenerator(trainset='trainval')
# get data reader
reader = create_fn(batch_size=100,
subset=subset,
version=FLAGS.test_version)
if checkpoint_path is None:
ckpt_dir = FLAGS.checkpoint_dir % (FLAGS.version, FLAGS.model_type)
# ckpt_dir = '/import/vision-ephemeral/fl302/models/v2_kpvaq_VAQ-RL/'
ckpt = tf.train.get_checkpoint_state(ckpt_dir)
checkpoint_path = ckpt.model_checkpoint_path
# Build model
g = tf.Graph()
with g.as_default():
# Build the model.
model = model_fn(model_config, 'beam')
model.build()
# Restore from checkpoint
restorer = Restorer(g)
sess = tf.Session()
restorer.restore(sess, checkpoint_path)
num_batches = reader.num_batches
print('Running beam search inference...')
results = []
for i in range(num_batches):
outputs = reader.get_test_batch()
# inference
quest_ids, image_ids = outputs[-2:]
scores, pathes = model.greedy_inference(outputs[:-2], sess)
scores, pathes = post_process_prediction(scores, pathes)
question = to_sentence.index_to_question(pathes[0])
print('%d/%d: %s' % (i, num_batches, question))
for quest_id, image_id, path in zip(quest_ids, image_ids, pathes):
sentence = to_sentence.index_to_question(path)
res_i = {
'image_id': int(image_id),
'question_id': int(quest_id),
'question': sentence
}
results.append(res_i)
save_json(res_file, results)
return res_file
def save_data(self):
save_payload = {
"trades": self.swaps,
"locks": self.locks,
"history": self.history,
"addresses": self.addresses
}
save_json(self.get_path(), save_payload)
def save(self):
if self.num_call % self.save_interval == 0:
print('Saving VQA replay buffers')
t = time()
sv_file = os.path.join(self.sv_dir, 'vqa_replay.json')
save_json(sv_file, {'num_call': self.num_call,
'memory': self.memory})
print('File %s saved to disk, total time: %0.2fs' % (sv_file, time() - t))
def process(method, inf_type='rand'):
if inf_type == 'rand':
res_file = 'result/tmp_bs_RL2_final_%s.json' % method
else:
res_file = 'result/tmp_bs_RL2_final_%s_BEAM.json' % method
if os.path.exists(res_file):
print('File %s already exist, skipped' % res_file)
return
# cands = load_results()
model = _TYPE2Model[method]()
mc_ctx = MultiChoiceQuestionManger(subset='val')
task_data = load_lm_outputs(method, inf_type)
belief_sets = {}
t = time()
num = len(task_data)
for i, ans_key in enumerate(task_data.keys()):
# time it
avg_time = (time() - t)
print('%d/%d (%0.2f sec/sample)' % (i, num, avg_time))
t = time()
# extract basis info
cands = task_data[ans_key]
quest_id = cands[0]['question_id']
# gt_answer = mc_ctx.get_gt_answer(quest_id)
image_id = mc_ctx.get_image_id(quest_id)
image = mc_ctx.get_image_file(quest_id)
# process
gt_question = mc_ctx.get_question(quest_id)
i_scores, i_questions = [], []
for item in cands:
target = item['question']
pred_ans, vqa_score = model.get_score(image_id, target)
# inset check
is_valid = compare_answer(pred_ans, ans_key)
if not is_valid:
continue
i_questions.append(target)
i_scores.append([float(vqa_score), item['score']])
print('%d/%d' % (len(i_questions), len(cands)))
bs_i = {
'image': image,
'image_id': image_id,
'question': gt_question,
'answer': ans_key,
'belief_sets': i_questions,
'belief_strength': i_scores
}
belief_sets[ans_key] = bs_i
save_json(res_file, belief_sets)
def vaq_decoding_greedy(checkpoint_path=None, subset='kpval'):
model_config = ModelConfig()
res_file = 'result/quest_vaq_greedy_%s.json' % FLAGS.model_type.upper()
# Get model
model_fn = get_model_creation_fn(FLAGS.model_type)
create_fn = create_reader(FLAGS.model_type, phase='test')
# Create the vocabulary.
to_sentence = SentenceGenerator(trainset='trainval')
# build data reader
reader = create_fn(batch_size=32, subset=subset)
if checkpoint_path is None:
ckpt = tf.train.get_checkpoint_state(FLAGS.checkpoint_dir %
FLAGS.model_type)
checkpoint_path = ckpt.model_checkpoint_path
g = tf.Graph()
with g.as_default():
# Build the model.
model = model_fn(model_config, 'greedy')
model.build()
saver = tf.train.Saver()
sess = tf.Session()
tf.logging.info('Restore from model %s' %
os.path.basename(checkpoint_path))
saver.restore(sess, checkpoint_path)
num_batches = reader.num_batches
print('Running greedy inference...')
results = []
for i in range(num_batches):
outputs = reader.get_test_batch()
# inference
quest_ids, image_ids = outputs[-2:]
scores, pathes = model.greedy_inference(outputs[:-2], sess)
scores, pathes = post_process_prediction(scores, pathes)
question = to_sentence.index_to_question(pathes[0])
print('%d/%d: %s' % (i, num_batches, question))
for quest_id, image_id, path in zip(quest_ids, image_ids, pathes):
sentence = to_sentence.index_to_question(path)
res_i = {
'image_id': int(image_id),
'question_id': int(quest_id),
'question': sentence
}
results.append(res_i)
save_json(res_file, results)
return res_file
def make_dummpy_result_file(pop_q):
from util import load_json, save_json
results = load_json('result/quest_vaq_greedy_VAQ-SAT_kptest.json')
for item in results:
item['image_id'] = int(item['image_id'])
item['question_id'] = int(item['question_id'])
item['question'] = pop_q
sv_file = 'result/tmp_pop_q_kptest.json'
save_json(sv_file, results)
return sv_file
def save_framesummary(self, frame):
data = {}
data['id'] = frame.id
data['name'] = frame.name
data['topic_ids'] = frame.topic_ids
data['bins'] = [(id, bin_name_from_id(id)) for id in frame.bin_ids]
data['counts'] = frame.counts
data['heatmapGrid'] = frame.heatmap.grid
data['heatmapWeights'] = self.get_frame_heatmap_weights(frame)
file_path = join(self.target_dir, 'framesummary', str(frame.id))
save_json(file_path, data, indent=0)
def save_topicframe(self, frame, topic_frame):
data = {}
data['topicId'] = topic_frame.topic_id
data['frameId'] = topic_frame.frame_id
data['tokenList'] = topic_frame.token_list
data['tokenWeights'] = self.get_token_weights(frame, topic_frame)
data['heatmapWeights'] = self.get_topicframe_heatmap_weights(
frame, topic_frame)
data['counts'] = topic_frame.counts
file_path = join(self.target_dir, 'topicframe',
str(topic_frame.frame_id), str(topic_frame.topic_id))
save_json(file_path, data, indent=0)
def get_authors(): files = util.listdir(AUTHOR_FOLDER) util.mkdir(AUTHOR_CRALWED_FOLDER) for file_name in files: save_path = os.path.join(AUTHOR_CRALWED_FOLDER, file_name) if util.exists(save_path): continue data = util.load_json(os.path.join(AUTHOR_FOLDER, file_name)) html = util.get_page(data['url']) full_name = get_full_name(html) data['name'] = full_name print data['short'], full_name data['links'] = get_links(data['short'], html) util.save_json(save_path, data)
def download_missing_paper_pdfs():
if not os.path.isdir(cfg.paths['pdfs-dir']):
os.makedirs(cfg.paths['pdfs-dir'])
metas = util.load_json(cfg.paths['papers-metadata'])
metas = util.parallelize(download_paper_pdf_if_needed, metas, N_THREADS)
util.save_json(cfg.paths['papers-metadata'], metas)
print('\n----')
print('saved updated papers metadata to "{}"'.format(
cfg.paths['papers-metadata']))
n = sum(int(os.path.isfile(m['pdf-path'])) for m in metas)
print('{}/{} ({:.3f}%) items with pdfs'.format(n, len(metas),
100 * n / len(metas)))
def get_authors():
files = util.listdir(AUTHOR_FOLDER)
util.mkdir(AUTHOR_CRALWED_FOLDER)
for file_name in files:
save_path = os.path.join(AUTHOR_CRALWED_FOLDER, file_name)
if util.exists(save_path):
continue
data = util.load_json(os.path.join(AUTHOR_FOLDER, file_name))
html = util.get_page(data['url'])
full_name = get_full_name(html)
data['name'] = full_name
print data['short'], full_name
data['links'] = get_links(data['short'], html)
util.save_json(save_path, data)
def build_question_set():
sv_file = 'data/kprestval_pos_tags.json'
st = StanfordPOSTagger('english-bidirectional-distsim.tagger')
meta = load_and_process_metadata('val')
images = split_data_by_seed(meta, 'kprestval')
num = len(images)
pos_tags_dict = {}
for i, info in enumerate(images):
question_id = info.question_id
question = info.question.lower()
_pos_tags = st.tag(word_tokenize(question))
pos_tags_dict[question_id] = _pos_tags
print('\nPOS TAGGER: %d/%d' % (i, num))
print(_pos_tags)
save_json(sv_file, {'pos_tags': pos_tags_dict})
def convert_vqa_annotations():
path_src = '../../data/VQA/Annotations/mscoco_val2014_annotations.json'
d = load_json(path_src)
anno = {}
for info in d['annotations']:
quest_id = int(info['question_id'])
tmp = {}
for _a in info['answers']:
a = _a['answer']
if a in tmp:
tmp[a] += 1.0
else:
tmp[a] = 1.0
anno[quest_id] = tmp
save_json(_ANNO_FILE, anno)
def get_conferences():
files = util.listdir(CONFERENCE_FOLDER)
util.mkdir(CONFERENCE_CRALWED_FOLDER)
cnt = 0
conf = util.load_json('conf_name.json')
for file_name in files:
save_path = os.path.join(CONFERENCE_CRALWED_FOLDER, file_name)
if util.exists(save_path):
continue
data = util.load_json(os.path.join(CONFERENCE_FOLDER, file_name))
if data['short'] not in conf.keys():
continue
html = util.get_page(data['url'])
subs = get_subs(data['short'], html)
data['name'] = conf[data['short']]
data['sub'] = {}
if len(subs) == 0:
data['sub']['#'] = get_publications(html)
util.save_json(save_path, data)
cnt += 1
def get_journals():
files = util.listdir(JOURNAL_FOLDER)
util.mkdir(JOURNAL_CRALWED_FOLDER)
cnt = 0
jour = util.load_json('jour_name.json')
for file_name in files:
save_path = os.path.join(JOURNAL_CRALWED_FOLDER, file_name)
if util.exists(save_path):
continue
data = util.load_json(os.path.join(JOURNAL_FOLDER, file_name))
if data['short'] not in jour.keys():
continue
html = util.get_page(data['url'])
subs = get_subs(data['short'], html)
data['name'] = jour[data['short']]
data['sub'] = {}
if len(subs) == 0:
data['sub']['#'] = get_publications(html)
util.save_json(save_path, data)
cnt += 1
print cnt, len(files), data['short']
def get_journals():
pos, cnt = 1, 0
util.mkdir(JOURNAL_FOLDER)
while True:
html = util.get_page(JOURNAL_URL + str(pos))
links = util.find_journals(html)
once_cnt = 0
for link in links:
if link[0] == '' or '?' in link[0]:
continue
data = {}
data['type'] = 'journal'
data['short'] = link[0]
data['name'] = link[1]
data['url'] = 'http://dblp.uni-trier.de/db/journals/' + data['short']
util.save_json(os.path.join(JOURNAL_FOLDER, util.hex_hash(data['short'])), data)
cnt += 1
once_cnt += 1
if once_cnt == 0:
break
pos += 100
print 'Journal', cnt
def get_conferences():
pos, cnt = 1, 0
util.mkdir(CONFERENCE_FOLDER)
while True:
html = util.get_page(CONFERENCE_URL + str(pos))
links = util.find_conferences(html)
once_cnt = 0
for link in links:
if link[0] == '' or '?' in link[0]:
continue
data = {}
data['type'] = 'conference'
data['short'] = link[0]
data['name'] = link[1]
data['url'] = 'http://dblp.uni-trier.de/db/conf/' + data['short']
util.save_json(os.path.join(CONFERENCE_FOLDER, util.hex_hash(data['short'])), data)
cnt += 1
once_cnt += 1
if once_cnt == 0:
break
pos += 100
print 'Conference', cnt
def get_authors():
pos, cnt = 545504, 0
util.mkdir(AUTHOR_FOLDER)
while True:
html = util.get_page(AUTHOR_URL + str(pos))
links = util.find_authors(html)
once_cnt = 0
for link in links:
if link[0] == '' or '?' in link[0]:
continue
data = {}
data['type'] = 'author'
data['short'] = link[0]
data['name'] = link[1]
data['url'] = 'http://dblp.uni-trier.de/pers/hd/a/' + data['short']
util.save_json(os.path.join(AUTHOR_FOLDER, util.hex_hash(data['short'])), data)
cnt += 1
once_cnt += 1
if once_cnt == 0:
break
pos += 300
print 'Author', pos, cnt
def get_journals():
files = util.listdir(JOURNAL_FOLDER)
util.mkdir(JOURNAL_CRALWED_FOLDER)
cnt = 0
jour = util.load_json('jour_name.json')
for file_name in files:
save_path = os.path.join(JOURNAL_CRALWED_FOLDER, file_name)
if util.exists(save_path):
continue
data = util.load_json(os.path.join(JOURNAL_FOLDER, file_name))
if data['short'] not in jour.keys():
continue
html = util.get_page(data['url'])
subs = get_subs(data['short'], html)
data['name'] = jour[data['short']]
data['sub'] = {}
for sub in subs:
html = util.get_page('http://dblp.uni-trier.de/db/journals/' + sub)
data['sub'][sub] = {}
data['sub'][sub]['pub'] = get_publications(html)
data['sub'][sub]['name'] = jour[sub]
cnt += 1
print cnt, len(files), data['short']
util.save_json(save_path, data)
ip_choices.remove(ip)
nodeStr = lb_node % (ip, node_port)
nodesStr += "%s\n " % nodeStr
if health_mon == "true":
hm = hm_req % (health_mon_type)
else:
hm = ""
lbStr = lb_post % (lb_name + "%i" % i, lb_port, lb_proto, vip_type, nodesStr, hm)
reqs.append(lbStr + "\n")
return reqs
if __name__ == "__main__":
prog = sys.argv[0]
if len(sys.argv) < 1:
usage(prog)
sys.exit()
ips = getIps("server_list.json")
request_list = []
configs = util.load_json("reqConfig.json")
for config in configs:
reqs = build_req(config, ips)
for req in reqs:
request_list.append(req)
random.shuffle(request_list)
util.save_json("requests.json", request_list)
JOURNAL_CRALWED_FOLDER = os.path.join('link', 'journal')
CONFERENCE_CRALWED_FOLDER = os.path.join('link', 'conference')
merged_data = {}
merged_data['jour'] = {}
merged_data['conf'] = {}
cnt = 0
files = util.listdir(JOURNAL_CRALWED_FOLDER)
for file_name in files:
data = util.load_json(os.path.join(JOURNAL_CRALWED_FOLDER, file_name))
short = data['short']
del data['short']
merged_data['jour'][short] = data
cnt += 1
print cnt, len(files)
cnt = 0
files = util.listdir(CONFERENCE_CRALWED_FOLDER)
for file_name in files:
data = util.load_json(os.path.join(CONFERENCE_CRALWED_FOLDER, file_name))
short = data['short']
del data['short']
merged_data['conf'][short] = data
cnt += 1
print cnt, len(files)
util.save_json('merged.json', merged_data)
