def getGt(n=1558):
gts = list()
for i in range(n):
for j in range(n):
if i == j:
gts.append("1")
else:
gts.append("0")
ut.writeList2Line(predictionPath, gtFilename, gts)
def getGt(n=1558):
gts = list()
for i in range(n):
for j in range(n):
if i == j:
gts.append("1")
else:
gts.append("0")
ut.writeList2Line(predictionPath, gtFilename, gts)
def nm(threshold=0.67):
instances = ut.readCommaLine2List(outputPath, featureNmFilename)
predictions = list()
count = 0
for instance in instances:
if (float(instance[1]) > threshold) or (float(instance[2])>threshold):
count += 1
predictions.append("1")
else:
predictions.append("0")
ut.writeList2Line(predPath, predictionNmFilename, predictions)
def writeMappingCandidates(): mappings = ut.readCommaLine2List(inputPath, "twitterMapping") candidates_google = list() candidates_twitter = list() for mapping in mappings: google_id = mapping[0] twitter_url = mapping[1] twitter_name = getTwitterUsername(twitter_url) if twitter_name != "": candidates_google.append(google_id) candidates_twitter.append(twitter_name) ut.writeList2Line(inputPath, "google/ids_mapping", candidates_google) ut.writeList2Line(inputPath, "twitter/names_mapping", candidates_twitter)
def ranking(n=1558, filename="ranking_origin_1558.txt"): scores = ut.readLine2List(predPath, filename) preds = list() for i in range(n): # print(i*n) scores_i = scores[i*n:(i+1)*n] max_index = max(enumerate(scores_i), key=lambda k: float(k[1]))[0] # print(max_index) preds_i = ["0"]*1558 preds_i[max_index] = "1" preds += preds_i ut.writeList2Line(predPath, predictionRankFilename, preds) return preds
def writeMappingCandidates():
mappings = ut.readCommaLine2List(inputPath, "twitterMapping")
candidates_google = list()
candidates_twitter = list()
for mapping in mappings:
google_id = mapping[0]
twitter_url = mapping[1]
twitter_name = getTwitterUsername(twitter_url)
if twitter_name != "":
candidates_google.append(google_id)
candidates_twitter.append(twitter_name)
ut.writeList2Line(inputPath, "google/ids_mapping", candidates_google)
ut.writeList2Line(inputPath, "twitter/names_mapping", candidates_twitter)
def writeMissingGooglePosts():
ids = ut.readLine2List("../data/google/", "ids_mapping")
ids_parsed = list()
ids_errors = list()
for root, folder, filenames in os.walk("../data/google/wall"):
ids_parsed = filenames
ids_errors = list(set(ids)-set(ids_parsed))
for filename in filenames:
with open(os.path.join(root, filename), "r", errors="ignore") as fi:
try:
result = json.loads(fi.read())
if type(result) == dict:
ids_errors.append(filename)
except:
pass
ut.writeList2Line("../data/stat/", "google_ids_post_errors", ids_errors)
def writeMissingGooglePosts():
ids = ut.readLine2List("../data/google/", "ids_mapping")
ids_parsed = list()
ids_errors = list()
for root, folder, filenames in os.walk("../data/google/wall"):
ids_parsed = filenames
ids_errors = list(set(ids) - set(ids_parsed))
for filename in filenames:
with open(os.path.join(root, filename), "r",
errors="ignore") as fi:
try:
result = json.loads(fi.read())
if type(result) == dict:
ids_errors.append(filename)
except:
pass
ut.writeList2Line("../data/stat/", "google_ids_post_errors", ids_errors)
def oneMapping(scores, outputFilename=predictionRankConstraintFilename, n=1558):
users1 = list()
users2 = dict()
predictions = list()
results = list()
# init
for i in range(n):
users2[i] = {"active": 0, "user": 0, "index": 0, "score": 0}
for i in range(n):
scores_i = scores[n*i:n*(i+1)]
scores_i_sorted = sorted(enumerate(scores_i), key=lambda k: k[1], reverse=True)
users1.append(scores_i_sorted)
# choose one mapping
for i in range(n):
oneMappingRecur(users1, users2, i, 0)
results = sorted([(v["user"], k) for k, v in users2.items()], key=operator.itemgetter(0))
for pair in results:
user2 = pair[1]
predictions_i = ["0"]*n
predictions_i[user2] = "1"
predictions+=predictions_i
ut.writeList2Line(predPath, outputFilename, predictions)
def reviseIdFile():
ids_visited = ut.readLine2List(snFolder, idsVisitedFileName+"2")
ids_saw = ut.readLine2List(snFolder, idsSawFileName)
loss = ut.readLine2List(snFolder, "tmp_ids")
# revise id file duplicate problem
g=nx.Graph()
dup = list()
num = list()
for i in range(len(ids_saw)):
id = ids_saw[i]
try:
g.node[id]
dup.append(id)
num.append(i)
except:
g.add_node(id)
print(len(dup))
for i in range(len(num)-1, -1, -1):
pos = num[i]
del ids_saw[pos]
for l in loss:
ids_saw.append(l)
ut.writeList2Line("../data/google/", "ids_saw2", ids_saw)
def reviseIdFile():
ids_visited = ut.readLine2List(snFolder, idsVisitedFileName + "2")
ids_saw = ut.readLine2List(snFolder, idsSawFileName)
loss = ut.readLine2List(snFolder, "tmp_ids")
# revise id file duplicate problem
g = nx.Graph()
dup = list()
num = list()
for i in range(len(ids_saw)):
id = ids_saw[i]
try:
g.node[id]
dup.append(id)
num.append(i)
except:
g.add_node(id)
print(len(dup))
for i in range(len(num) - 1, -1, -1):
pos = num[i]
del ids_saw[pos]
for l in loss:
ids_saw.append(l)
ut.writeList2Line("../data/google/", "ids_saw2", ids_saw)
