def main():
if len(sys.argv) < 3:
sys.stderr.write('Usage: %s outdir textfile1 textfile2 ...\n'
% sys.argv[0])
sys.exit(1)
outdir = sys.argv[1]
tfdb = DB()
if not tfdb.open(os.path.join(outdir, 'tf.kch'),
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE):
sys.stderr.write('cannot open tfdb: %s\n' % str(tfdb.error))
sys.exit(1)
dfdb = DB()
if not dfdb.open(os.path.join(outdir, 'df.kch'),
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE):
sys.stderr.write('cannot open dfdb: %s\n' % str(dfdb.error))
sys.exit(1)
tfidfdb = DB()
if not tfidfdb.open(os.path.join(outdir, 'tfidf.kch'),
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE):
sys.stderr.write('cannot open tfidfdb: %s\n' % str(tfidfdb.error))
sys.exit(1)
print 'Count words ...'
for i in range(len(sys.argv)-2):
filename = sys.argv[i+2]
print '(%d/%d) %s' % (i+1, len(sys.argv)-2, filename)
count_words(tfdb, dfdb, filename)
print 'Calculate TFIDF ...'
save_tfidf(tfdb, dfdb, tfidfdb)
tfdb.close()
dfdb.close()
tfidfdb.close()
def main():
if len(sys.argv) < 3:
sys.stderr.write('Usage: %s outdir textfile1 textfile2 ...\n' %
sys.argv[0])
sys.exit(1)
outdir = sys.argv[1]
tfdb = DB()
if not tfdb.open(os.path.join(outdir, 'tf.kch'),
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE):
sys.stderr.write('cannot open tfdb: %s\n' % str(tfdb.error))
sys.exit(1)
dfdb = DB()
if not dfdb.open(os.path.join(outdir, 'df.kch'),
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE):
sys.stderr.write('cannot open dfdb: %s\n' % str(dfdb.error))
sys.exit(1)
tfidfdb = DB()
if not tfidfdb.open(os.path.join(outdir, 'tfidf.kch'),
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE):
sys.stderr.write('cannot open tfidfdb: %s\n' % str(tfidfdb.error))
sys.exit(1)
print 'Count words ...'
for i in range(len(sys.argv) - 2):
filename = sys.argv[i + 2]
print '(%d/%d) %s' % (i + 1, len(sys.argv) - 2, filename)
count_words(tfdb, dfdb, filename)
print 'Calculate TFIDF ...'
save_tfidf(tfdb, dfdb, tfidfdb)
tfdb.close()
dfdb.close()
tfidfdb.close()
def open_db(self, name):
db = DB()
db.open(
join(self.path, "%s.kch"%name),
DB.OWRITER | DB.OCREATE
)
self.opendb.append(db)
return db
class KDB:
def __init__(self, path, truncate=False):
"""
Open a new connection to a database using the Kyoto Cabinet engine.
Args:
* path (str): Path to database.
Kwargs:
* truncate (bool, False): If database should be truncated before opening.
"""
self.db = DB()
self.batchsize = 1000
self.batch = {}
dbparams = '.kct#apow=0#bnum=10000000#msiz=' + str(2 << 30)
if truncate:
result = self.db.open(path + dbparams,
DB.OWRITER | DB.OCREATE | DB.OTRUNCATE)
else:
result = self.db.open(path + dbparams, DB.OWRITER)
if not result:
raise PathError('DNA outdb open error: %s ' % self.db.error())
exit(1)
def get(self, key):
"""
Retrieve the item with the given `key`.
"""
return self.db.get(key)
def put(self, key, val):
"""
Put `val` at `key`.
Note that disk writing is done in batches, so be sure to call `close` or `flush` to make sure that values are put into the store.
"""
self.batch[key] = val
if len(self.batch) >= self.batchsize:
self.flush()
def flush(self):
"""
Write `put` calls to database.
"""
self.db.set_bulk(self.batch)
self.batch = {}
def close(self):
"""
Flush the database and delete the connection to it.
"""
self.flush()
del self.db
def dbOpen(name):
db = DB()
if self.create:
# if not db.open(abspath(self.path) + '/' + name + ".kch",
# DB.OWRITER | DB.OCREATE | DB.OAUTOSYNC | DB.OAUTOTRAN):
if not db.open(abspath(self.path) + "/" + name + ".kch", DB.OWRITER | DB.OCREATE):
raise IOError("open error: " + str(db.error()))
return db
else:
# if not db.open(abspath(self.path) + '/' + name + ".kch",
# DB.OWRITER | DB.OAUTOSYNC | DB.OAUTOTRAN):
if not db.open(abspath(self.path) + "/" + name + ".kch", DB.OWRITER):
raise IOError("open error: " + str(db.error()))
return db
def dbOpen(name):
db = DB()
dbpathname = abspath(self.path) + '/' + name + ".kch"
if self.create:
# if not db.open(abspath(self.path) + '/' + name + ".kch",
# DB.OWRITER | DB.OCREATE | DB.OAUTOSYNC | DB.OAUTOTRAN):
if not db.open(dbpathname, DB.OWRITER | DB.OCREATE):
raise IOError("open error: %s %s" % (dbpathname, str(db.error()))) #pragma: NO COVER
return db
else:
# if not db.open(abspath(self.path) + '/' + name + ".kch",
# DB.OWRITER | DB.OAUTOSYNC | DB.OAUTOTRAN):
if not db.open(dbpathname, DB.OWRITER): #pragma: NO COVER
raise IOError("open error: %s %s" % (dbpathname, str(db.error()))) #pragma: NO COVER
return db
def get_items(item_filter, db_file, page=0):
item_iter = 0
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while len(items) < FILTER_MAX:
rec = cur.get(False)
if not rec:
break
if item_iter != (FILTER_MAX * page):
if item_filter(rec):
item_iter = item_iter + 1
cur.step_back()
continue
if item_filter(rec):
items.append(rec)
cur.step_back()
cur.disable()
db.close()
sorted_items = sorted(items, key=get_key, reverse=True)
sorted_items_for_viewing = [loads(item[1]) for item in sorted_items]
for item in sorted_items_for_viewing:
if item['title'] is None or item['title'] == "":
item['title'] = item['url']
return sorted_items_for_viewing
class DbKyoto(object):
def __init__(self, db_file):
#from train import TAG2ID, WORD2ID#, BAYES_RANK
#self.ider = WORD2ID
self.db = DB()
self.db_file = db_file
print path.join(KYOTO_DB_PATH,self.db_file)
if not self.db.open(path.join(KYOTO_DB_PATH,self.db_file), DB.OWRITER | DB.OCREATE):
print >>sys.stderr, "open error: " + str(self.db.error())
def set(self,entry):
key = entry[0]
result_array = convert2array(entry[1]).tostring()
if not self.db.set(key,result_array):
print key
print result_array
print >>sys.stderr, "open error: " + str(self.db.error())
def get(self,key):
value = self.db.get(key)
if value:
result = array('L')
result.fromstring(value)
return convert2dict(result)
else:
#print >>sys.stderr, self.ider.get_word_by_id(key)
#print key
#print >>sys.stderr, "%s error: "%key + str(self.db.error())
pass
class DbKyoto(object):
def __init__(self, db_file):
#from train import TAG2ID, WORD2ID#, BAYES_RANK
#self.ider = WORD2ID
self.db = DB()
self.db_file = db_file
print path.join(KYOTO_DB_PATH, self.db_file)
if not self.db.open(path.join(KYOTO_DB_PATH, self.db_file),
DB.OWRITER | DB.OCREATE):
print >> sys.stderr, "open error: " + str(self.db.error())
def set(self, entry):
key = entry[0]
result_array = convert2array(entry[1]).tostring()
if not self.db.set(key, result_array):
print key
print result_array
print >> sys.stderr, "open error: " + str(self.db.error())
def get(self, key):
value = self.db.get(key)
if value:
result = array('L')
result.fromstring(value)
return convert2dict(result)
else:
#print >>sys.stderr, self.ider.get_word_by_id(key)
#print key
#print >>sys.stderr, "%s error: "%key + str(self.db.error())
pass
def aggregate_by_hour(db_file):
# Initialize the dict with each hour
hours = {key: 0 for key in range(0,24)}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
unix = float(loaded['created_at'])
time = datetime.fromtimestamp(unix)
hours[time.hour] = hours[time.hour] + 1
cur.step_back()
cur.disable()
db.close()
hours = [{'name': "{}:00".format(key), 'data': [hours[key]]} for key in hours]
return hours
def get_items(item_filter, db_file, page=0):
item_iter = 0
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while len(items) < FILTER_MAX:
rec = cur.get(False)
if not rec:
break
if item_iter != (FILTER_MAX * page):
if item_filter(rec):
item_iter = item_iter + 1
cur.step_back()
continue
if item_filter(rec):
items.append(rec)
cur.step_back()
cur.disable()
db.close()
sorted_items = sorted(items, key=get_key, reverse=True)
sorted_items_for_viewing = [loads(item[1]) for item in sorted_items]
for item in sorted_items_for_viewing:
if item['title'] is None or item['title'] == "":
item['title'] = item['url']
return sorted_items_for_viewing
def get_post_num(post_num, db_file):
item = None
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump()
i = 0
while True:
rec = cur.get(False)
if not rec:
break
if i == post_num:
item = rec
cur.step()
i = i + 1
cur.disable()
db.close()
if item is not None:
return loads(item[1])
return dict()
class DbKyoto(object):
def __init__(self, db_file):
self.db = DB()
self.db_file = db_file
if not self.db.open(db_file, DB.OWRITER | DB.OCREATE):
print >> sys.stderr, 'open error: ' + str(self.db.error())
def set(self, txt, po_id):
feature_list = feature_md5(txt)
for feature in feature_list:
key = feature
entry = self.get(key)
if not entry:
val = array('L', [po_id])
if not self.db.set(key, val.tostring()):
print >> sys.stderr, 'open error: ' + str(self.db.error())
else:
val = array('L')
val.fromstring(entry)
if po_id not in val:
val.append(po_id)
self.db.set(key, val.tostring())
return val
def get(self, key):
po_id = self.db.get(key)
result = array('L')
if po_id:
result.fromstring(po_id)
return result
def get_post_num(post_num, db_file):
item = None
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump()
i = 0
while True:
rec = cur.get(False)
if not rec:
break
if i == post_num:
item = rec
cur.step()
i = i + 1
cur.disable()
db.close()
if item is not None:
return loads(item[1])
return dict()
def aggregate_by_hour(db_file):
# Initialize the dict with each hour
hours = {key: 0 for key in range(0, 24)}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
unix = float(loaded['created_at'])
time = datetime.fromtimestamp(unix)
hours[time.hour] = hours[time.hour] + 1
cur.step_back()
cur.disable()
db.close()
hours = [{
'name': "{}:00".format(key),
'data': [hours[key]]
} for key in hours]
return hours
class DbKyoto(object):
def __init__(self, db_file):
self.db = DB()
self.db_file = db_file
if not self.db.open(db_file, DB.OWRITER | DB.OCREATE):
print >> sys.stderr, 'open error: ' + str(self.db.error())
def set(self, txt, po_id):
feature_list = feature_md5(txt)
for feature in feature_list:
key = feature
entry = self.get(key)
if not entry:
val = array('L', [po_id])
if not self.db.set(key, val.tostring()):
print >> sys.stderr, 'open error: ' + str(self.db.error())
else:
val = array('L')
val.fromstring(entry)
if po_id not in val:
val.append(po_id)
self.db.set(key, val.tostring())
return val
def get(self, key):
po_id = self.db.get(key)
result = array('L')
if po_id:
result.fromstring(po_id)
return result
def __load_blast_data(blast):
# Connect to kyoto db
db = DB()
if not db.open("/opt/gene2accession/gene2accession.kch", DB.OREADER):
raise Exception("Could not load gene2accession.kch: " + str(db.error()))
hits = {}
gi_num = re.compile('gi\|([0-9]+)')
for line in blast:
split_line = line.split('\t')
# Important data
evalue = float(split_line[10])
gi_nums = gi_num.findall(split_line[12])
genome_ids = [db.get(x) for x in gi_nums if db.get(x) is not None]
# Thanks to Peter's parser, the gi list and org list are the same
# length (the first gi column is also the first gi in the "master" gi
# column)
for org in genome_ids:
if org in hits:
hits[org].append(evalue)
else:
hits[org] = [evalue]
db.close()
return hits
def top_things(db_file):
urls = {}
people = {}
graph = {}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database. (Top things)"
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded_rec = loads(rec[1])
split = get_domain(loaded_rec)
if urls.get(split, False) == False:
urls[split] = 1
else:
urls[split] = urls[split] + 1
person = loaded_rec['person']
if people.get(person, False) == False:
people[person] = 1
else:
people[person] = people[person] + 1
if split is not None and split is not "" and \
person is not None and person is not "":
# Build a crazy relational graph out of my nosql data
if graph.get(split, False) == False:
graph[split] = {
"is_person": False,
"data": [person],
"linked_to_count": 1
}
elif person not in graph[split]:
graph[split]["data"].append(person)
graph[split][
"linked_to_count"] = graph[split]["linked_to_count"] + 1
if graph.get(person, False) == False:
graph[person] = {"is_person": True, "data": [split]}
elif split not in graph[person]:
graph[person]["data"].append(split)
cur.step_back()
cur.disable()
db.close()
def get_one(x):
return x[1]
return (sorted(urls.items(), key=get_one, reverse=True),
sorted(people.items(), key=get_one, reverse=True), graph)
def decorated_function(*args, **kwargs):
# Debug
if not current_app.config['CACHE']:
return f(*args, **kwargs)
db = DB()
db.open("/tmp/page_cache.kch")
res = None
fancy = hash("{}{}{}".format(db_meta_info()['count'], request.url, f.func_name))
res = db.get(fancy)
if not res:
res = f(*args, **kwargs)
db.set(fancy, res)
db.close()
return res
def open(self, filename):
db = DB()
# self.db=DB(DB.GEXCEPTIONAL)
if not db.open(filename, DB.OWRITER | DB.OCREATE | DB.ONOLOCK):
raise IOError("open error: '" + str(self.db.error()) + "' on file:"
+ filename)
return db
def dbOpen(name):
db = DB()
dbpathname = abspath(self.path) + '/' + name + ".kch"
if self.create:
# if not db.open(abspath(self.path) + '/' + name + ".kch",
# DB.OWRITER | DB.OCREATE | DB.OAUTOSYNC | DB.OAUTOTRAN):
if not db.open(dbpathname, DB.OWRITER | DB.OCREATE):
raise IOError("open error: %s %s" % (
dbpathname, str(db.error()))) # pragma: NO COVER
return db
else:
# if not db.open(abspath(self.path) + '/' + name + ".kch",
# DB.OWRITER | DB.OAUTOSYNC | DB.OAUTOTRAN):
if not db.open(dbpathname, DB.OWRITER): # pragma: NO COVER
raise IOError("open error: %s %s" % (
dbpathname, str(db.error()))) # pragma: NO COVER
return db
def __init__(self, dbpath, readonly=False):
self.readonly = readonly
mode = KDB.OREADER if readonly else KDB.OREADER | KDB.OWRITER | KDB.OCREATE
db = KDB()
if not db.open(dbpath, mode):
raise IOError("kyotocabinet.DB().open(%s, %o): %s" %
(dbpath, mode, db.error()))
self.db = db
self.dbpath = dbpath
def insert_item(url, person, db_file, submitted_title=''):
mimetype = "application/json"
db = DB()
if not db.open("{0}".format(db_file), DB.OWRITER | DB.OCREATE):
response = {}
response[
'What happened?'] = "Couldn't open the damn database. Error: {0}".format(
db.error())
return Response(dumps(response), mimetype=mimetype)
if is_url_in_db(db, url):
return Response('{"What happened?": "Someone '\
'tried to submit a duplicate URL."}',
mimetype=mimetype)
title = url
summary = "~?~"
try:
thing = urlopen(url, timeout=10)
soup = BeautifulSoup(thing)
title = soup.title.string
# Do some dumb summarizing if we can
def concat(a, v):
return a + " " + v.strip()
visible_stuff = filter(visible, soup.findAll(text=True))
summary = reduce(concat, visible_stuff, "")[:900] + "..."
except:
pass
#return Response('{"What happened?": '\
# 'I dunno bs4 messed up somehow."}',
# mimetype=mimetype)
created_at = int(mktime(datetime.now().utctimetuple()))
is_image = url.lower().endswith(("jpg", "jpeg", "gif", "png"))
thumbnail = gen_thumbnail_for_url(url, str(created_at))
record = {
"created_at": created_at,
"title": title,
"url": url,
"person": person,
"summary": summary,
"person_color": PERSON_COLORS[random.randint(0,
len(PERSON_COLORS) - 1)],
"is_image": is_image,
"thumbnail": thumbnail,
"comment": submitted_title
}
db.set(created_at, dumps(record))
db.close()
return Response('{"What happened?": "MUDADA"}', mimetype=mimetype)
def insert_item(url, person, db_file, submitted_title=''):
mimetype = "application/json"
db = DB()
if not db.open("{0}".format(db_file),
DB.OWRITER | DB.OCREATE):
response = {}
response['What happened?'] = "Couldn't open the damn database. Error: {0}".format(db.error())
return Response(dumps(response), mimetype=mimetype)
if is_url_in_db(db, url):
return Response('{"What happened?": "Someone '\
'tried to submit a duplicate URL."}',
mimetype=mimetype)
title = url
summary = "~?~"
try:
thing = urlopen(url, timeout=10)
soup = BeautifulSoup(thing)
title = ''
if len(submitted_title) > 0:
title = submitted_title
else:
title = soup.title.string
# Do some dumb summarizing if we can
func = lambda a,v: a + " " + v.strip()
visible_stuff = filter(visible, soup.findAll(text=True))
summary = reduce(func, visible_stuff, "")[:900] + "..."
except:
pass
#return Response('{"What happened?": '\
# 'I dunno bs4 messed up somehow."}',
# mimetype=mimetype)
created_at = int(mktime(datetime.now().utctimetuple()))
is_image = url.lower().endswith(("jpg", "jpeg", "gif", "png"))
thumbnail = gen_thumbnail_for_url(url, str(created_at))
record = {
"created_at": created_at,
"title": title,
"url": url,
"person": person,
"summary": summary,
"person_color": PERSON_COLORS[random.randint(0, len(PERSON_COLORS)-1)],
"is_image": is_image,
"thumbnail": thumbnail
}
db.set(created_at, dumps(record))
db.close()
return Response('{"What happened?": "MUDADA"}',
mimetype=mimetype)
def top_things(db_file):
urls = {}
people = {}
graph = {}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database. (Top things)"
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded_rec = loads(rec[1])
split = get_domain(loaded_rec)
if urls.get(split, False) == False:
urls[split] = 1
else:
urls[split] = urls[split] + 1
person = loaded_rec['person']
if people.get(person, False) == False:
people[person] = 1
else:
people[person] = people[person] + 1
if split is not None and split is not "" and \
person is not None and person is not "":
# Build a crazy relational graph out of my nosql data
if graph.get(split, False) == False:
graph[split] = {"is_person": False, "data": [person], "linked_to_count": 1}
elif person not in graph[split]:
graph[split]["data"].append(person)
graph[split]["linked_to_count"] = graph[split]["linked_to_count"] + 1
if graph.get(person, False) == False:
graph[person] = {"is_person": True, "data": [split]}
elif split not in graph[person]:
graph[person]["data"].append(split)
cur.step_back()
cur.disable()
db.close()
def get_one(x):
return x[1]
return (sorted(urls.items(), key=get_one, reverse=True),
sorted(people.items(), key=get_one, reverse=True),
graph)
def db_meta_info():
meta = {}
db = DB()
db_file = current_app.config['DB_FILE']
if not db.open("{0}".format(db_file), DB.OREADER):
print "Could not open database (meta info)."
meta["size"] = db.size()
meta["count"] = db.count()
db.close()
return meta
def get_post_by_date(key, db_file):
item = None
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
item = db.get(key)
db.close()
if item is not None:
return loads(item)
return dict()
def get_post_by_date(key, db_file):
item = None
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
item = db.get(key)
db.close()
if item is not None:
return loads(item)
return dict()
def db_meta_info():
meta = {}
db = DB()
db_file = current_app.config['DB_FILE']
if not db.open("{0}".format(db_file), DB.OREADER):
print "Could not open database (meta info)."
meta["size"] = db.size()
meta["count"] = db.count()
db.close()
return meta
def main():
db = DB()
if not db.open("bayes.kch", DB.OWRITER | DB.OCREATE):
return
with open("word_tf.txt") as word_tf:
for line in word_tf:
line = line.strip()
word, bayes_list = loads(line)
print word
if bayes_list:
ar = array('I')
ar.fromlist(lineiter(bayes_list))
db[word] = ar.tostring()
def main():
db = DB()
if not db.open("bayes.kch", DB.OWRITER | DB.OCREATE):
return
with open("word_tf.txt") as word_tf:
for line in word_tf:
line = line.strip()
word, bayes_list = loads(line)
print word
if bayes_list:
ar = array('I')
ar.fromlist(lineiter(bayes_list))
db[word] = ar.tostring()
def purge(domain, genid):
if request.remote_addr not in settings.ALLOW:
return text_response("Not permitted.\n", 403)
db = DB()
if not db.open(settings.GENID_DATABASE, DB.OWRITER | DB.OCREATE):
return text_response("Failed to purge: cannot open database.\n", 501)
set_ok = db.set(domain, genid)
db.close()
if not set_ok:
return text_response("Failed to purge: cannot set genid.\n", 501)
else:
return text_response("Purged <%s>\n" % (domain,))
def get_items_last_X_days(db_file, X, munge=True):
dates = {}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
X_days_ago = datetime.now() - timedelta(days=X)
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
unix = float(loaded['created_at'])
time = datetime.fromtimestamp(unix)
if time > X_days_ago:
if munge:
date_obj = date(year=time.year, month=time.month, day=time.day)
else:
date_obj = time
# Javascript expects Date.UTC to spit out dates of a certain
# length.
day_unix = int(mktime(date_obj.timetuple())) * 1000
if dates.get(day_unix, False) == False:
dates[day_unix] = {loaded["person"]: 1}
else:
relevant_dict = dates[day_unix]
if relevant_dict.get(loaded["person"], False) == False:
relevant_dict[loaded["person"]] = 1
else:
relevant_dict[
loaded["person"]] = relevant_dict[loaded["person"]] + 1
else:
break
cur.step_back()
cur.disable()
db.close()
return dates
def get_items_last_X_days(db_file, X, munge=True):
dates = {}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
X_days_ago = datetime.now() - timedelta(days=X)
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
unix = float(loaded['created_at'])
time = datetime.fromtimestamp(unix)
if time > X_days_ago:
if munge:
date_obj = date(year=time.year, month=time.month, day=time.day)
else:
date_obj = time
# Javascript expects Date.UTC to spit out dates of a certain
# length.
day_unix = int(mktime(date_obj.timetuple()))*1000
if dates.get(day_unix, False) == False:
dates[day_unix] = {loaded["person"]: 1}
else:
relevant_dict = dates[day_unix]
if relevant_dict.get(loaded["person"], False) == False:
relevant_dict[loaded["person"]] = 1
else:
relevant_dict[loaded["person"]] = relevant_dict[loaded["person"]] + 1
else:
break;
cur.step_back()
cur.disable()
db.close()
return dates
def get_last_items(db_file, pages=1):
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while len(items) < (pages * FILTER_MAX):
rec = cur.get(False)
if not rec:
break
items.append(rec)
cur.step_back()
cur.disable()
db.close()
return items
def get_last_items(db_file, pages=1):
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while len(items) < (pages * FILTER_MAX):
rec = cur.get(False)
if not rec:
break
items.append(rec)
cur.step_back()
cur.disable()
db.close()
return items
def get_all_items(db_file):
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump()
while True:
rec = cur.get(False)
if not rec:
break
items.append(rec)
cur.step()
cur.disable()
db.close()
sorted_items_for_viewing = [loads(item[1]) for item in items]
return sorted_items_for_viewing
def get_all_items(db_file):
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump()
while True:
rec = cur.get(False)
if not rec:
break
items.append(rec)
cur.step()
cur.disable()
db.close()
sorted_items_for_viewing = [loads(item[1]) for item in items]
return sorted_items_for_viewing
def get_page_count(item_filter = lambda x: True):
count = 0
db = DB()
db_file = current_app.config['DB_FILE']
if not db.open("{0}".format(db_file), DB.OREADER | DB.OWRITER | DB.OCREATE):
print "Could not open database (get_page_count). Error: {}".format(db.error())
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
if item_filter(rec):
count = count + 1
cur.step_back()
cur.disable()
db.close()
return count / FILTER_MAX
def gen_thumbnails(db_file):
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OWRITER):
sys.exit(1)
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
is_image = loaded["url"].lower().endswith(
("jpg", "jpeg", "gif", "png"))
if is_image:
print "Thumbnailing {}".format(loaded["url"])
loaded["is_image"] = True
try:
thumbnail = gen_thumbnail_for_url(loaded["url"], rec[0])
except IOError as e:
print "IOError: {}".format(e)
print "Save result: {}".format(cur.set_value(dumps(loaded)))
cur.step_back()
continue
if thumbnail:
loaded["thumbnail"] = thumbnail
print "Thumbnailed {}".format(loaded["url"])
print "Save result: {}".format(cur.set_value(dumps(loaded)))
cur.step_back()
cur.disable()
db.close()
return True
def gen_thumbnails(db_file):
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OWRITER):
sys.exit(1)
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
is_image = loaded["url"].lower().endswith(("jpg", "jpeg", "gif", "png"))
if is_image:
print "Thumbnailing {}".format(loaded["url"])
loaded["is_image"] = True
try:
thumbnail = gen_thumbnail_for_url(loaded["url"], rec[0])
except IOError as e:
print "IOError: {}".format(e)
print "Save result: {}".format(cur.set_value(dumps(loaded)))
cur.step_back()
continue
if thumbnail:
loaded["thumbnail"] = thumbnail
print "Thumbnailed {}".format(loaded["url"])
print "Save result: {}".format(cur.set_value(dumps(loaded)))
cur.step_back()
cur.disable()
db.close()
return True
def get_items_on_page(page, db_file):
item_iter = 0
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while len(items) < FILTER_MAX:
rec = cur.get(False)
if not rec:
break
if item_iter >= (FILTER_MAX * page):
items.append(rec)
item_iter = item_iter + 1
cur.step_back()
cur.disable()
db.close()
return items
def get_items_on_page(page, db_file):
item_iter = 0
items = []
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump_back()
while len(items) < FILTER_MAX:
rec = cur.get(False)
if not rec:
break
if item_iter >= (FILTER_MAX * page):
items.append(rec)
item_iter = item_iter + 1
cur.step_back()
cur.disable()
db.close()
return items
def main():
db_file = argv[1]
username = argv[2]
if not db_file and not username:
print "Need db_file and username."
return -1
db = DB()
if not db.open("{0}".format(db_file), DB.OWRITER):
print "Could not open database."
return -1
all_keys = []
cur = db.cursor()
cur.jump()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
if loaded["person"] == username:
all_keys.append(cur.get_key())
cur.step()
cur.disable()
print "Found {} records.".format(len(all_keys))
for key in all_keys:
print "Pending {}...".format(key)
if len(argv) > 3 and argv[3] == '--delete':
print "Removing {}...".format(key)
if not db.remove(key):
print "Could not remove key: {}".format(db.error())
db.close()
def main():
db_file = argv[1]
username = argv[2]
if not db_file and not username:
print "Need db_file and username."
return -1
db = DB()
if not db.open("{0}".format(db_file), DB.OWRITER):
print "Could not open database."
return -1
all_keys = []
cur = db.cursor()
cur.jump()
while True:
rec = cur.get(False)
if not rec:
break
loaded = loads(rec[1])
if loaded["person"] == username:
all_keys.append(cur.get_key())
cur.step()
cur.disable()
print "Found {} records.".format(len(all_keys))
for key in all_keys:
print "Pending {}...".format(key)
if len(argv) > 3 and argv[3] == '--delete':
print "Removing {}...".format(key)
if not db.remove(key):
print "Could not remove key: {}".format(db.error())
db.close()
def get_page_count(item_filter=lambda x: True):
count = 0
db = DB()
db_file = current_app.config['DB_FILE']
if not db.open("{0}".format(db_file),
DB.OREADER | DB.OWRITER | DB.OCREATE):
print "Could not open database (get_page_count). Error: {}".format(
db.error())
cur = db.cursor()
cur.jump_back()
while True:
rec = cur.get(False)
if not rec:
break
if item_filter(rec):
count = count + 1
cur.step_back()
cur.disable()
db.close()
return count / FILTER_MAX
class DataStorage(object):
"""
Parent class for RowData and KeyValueData.
"""
def __init__(self, filename, headers=None):
self.filename = filename
self.ext = os.path.splitext(filename)[1]
self.headers = headers
if os.path.exists(self.filename):
self.init_read()
else:
self.init_write()
def init_write(self):
self.mode = "write"
if self.ext == ".csv":
self._data_file = open(self.filename, "wb")
self._writer = csv.writer(self._data_file)
if self.headers:
self._writer.writerow(self.headers)
elif self.ext == ".json":
self._storage = {}
elif self.ext == ".kch":
from kyotocabinet import DB
self._storage = DB()
if not self._storage.open(self.filename, DB.OWRITER | DB.OCREATE):
msg = "Error opening kyotocabinet db: %s" % (
self._storage.error())
raise dexy.exceptions.UserFeedback(msg)
elif self.ext == ".sqlite3":
self.init_write_sqlite3()
else:
raise dexy.exceptions.UserFeedback("unsupported extension %s" %
self.ext)
def init_read(self):
self.mode = "read"
if self.ext == ".csv":
self._file = open(self.filename, "rb")
elif self.ext == ".json":
with open(self.filename, "rb") as f:
self._storage = json.load(f)
elif self.ext == ".kch":
from kyotocabinet import DB
self._storage = DB()
self._storage.open(self.filename, DB.OREADER)
elif self.ext == ".sqlite3":
import sqlite3
self._storage = sqlite3.connect(self.filename)
self._cursor = self._storage.cursor()
else:
raise dexy.exceptions.UserFeedback("unsupported extension %s" %
self.ext)
def save(self):
if self.ext == ".csv":
self._data_file.close()
elif self.ext == ".json":
with open(self.filename, "wb") as f:
import json
json.dump(self._storage, f)
elif self.ext == ".kch":
if not self._storage.close():
raise dexy.exceptions.UserFeedback(self._storage.error())
elif self.ext == ".sqlite3":
self._storage.commit()
self._cursor.close()
else:
raise dexy.exceptions.UserFeedback("unsupported extension %s" %
self.ext)
from kyotocabinet import DB
import sys,time
#format datetime
def format_datetime(timestamp):
return time.strftime('%Y.%m.%d @ %H:%M',time.localtime(timestamp))
#创建数据库对象
db=DB()
# open the database
if not db.open("../data/db.kch", DB.OWRITER | DB.OCREATE):
print >>sys.stderr, "open error: " + str(db.error())
cid='1'
#build for marks form
if not db.set("mark:"+cid+":markid", "1") or\
not db.set("mark:"+cid+":userid", "1") or\
not db.set("mark:"+cid+":boardid", "1") or\
not db.set("mark:"+cid+":fileid", "1") or\
not db.set("mark:"+cid+":description", "mark的内容描述") or\
not db.set("mark:"+cid+":content", "哇。。阳光暖暖的，好惬意的。") or\
not db.set("mark:"+cid+":ups", "100") or\
not db.set("mark:"+cid+":downs", "10") or\
not db.set("mark:"+cid+":hits", "110") or\
not db.set("mark:"+cid+":order", "1") or\
not db.set("maks:"+cid+":createdata", int(time.time())) or\
not db.set("maks:"+cid+":commentcount", "8") or\
def get_user_stats(username, db_file):
item = {
"username": username,
"aliases": [],
"total_posts": 0,
"domains": {},
"first_post_date": None,
"first_post_date_unix": None,
"most_recent_post": None,
"most_recent_post_unix": 0,
"average_posts_per_hour": 0.0,
"average_posts_per_day": 0.0,
"average_posts_per_week": 0.0
}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump()
while True:
rec = cur.get(False)
if not rec:
break
loaded_rec = loads(rec[1])
if loaded_rec['person'] != username:
cur.step()
continue
# Looks like this is a post by the user we're looking for
split = get_domain(loaded_rec)
if item['domains'].get(split, False) == False:
item['domains'][split] = 1
else:
item['domains'][split] = item['domains'][split] + 1
if item['first_post_date_unix'] is None:
item['first_post_date_unix'] = loaded_rec['created_at']
if item['most_recent_post_unix'] < loaded_rec['created_at']:
item['most_recent_post_unix'] = loaded_rec['created_at']
item['total_posts'] = item['total_posts'] + 1
cur.step()
cur.disable()
db.close()
# Clean up everything
first_time = None
if item['first_post_date_unix'] is not None:
unix = float(item['first_post_date_unix'])
first_time = datetime.fromtimestamp(unix)
item['first_post_date'] = first_time.isoformat()
recent_time = None
if item['most_recent_post_unix'] is not None:
unix = float(item['most_recent_post_unix'])
recent_time = datetime.fromtimestamp(unix)
item['most_recent_post'] = recent_time.isoformat()
if first_time and recent_time:
delta = recent_time - first_time
item['user_age_days'] = delta.days
item['user_age_seconds'] = delta.total_seconds()
item['average_posts_per_hour'] = item['total_posts'] / (delta.total_seconds() / 60.0)
item['average_posts_per_day'] = item['total_posts'] / (delta.total_seconds() / 60.0 / 24.0)
item['average_posts_per_week'] = item['total_posts'] / (delta.total_seconds() / 60.0 / 24.0 / 7.0)
return item
def convert2array(dict_value):
'''
>>> convert2array({1:0.1,2:0.3})
array('L', [1L, 429496729L, 2L, 1288490188L])
'''
result_list = []
for k, v in dict_value:
result_list.extend([k, int(v * MAX_INT)])
result = array('L', result_list)
return result
def convert2dict(array_l):
'''
>>> convert2dict([1L, 429496729L, 2L, 1288490188L])
{1:0.1,2:0.3}
'''
return [(array_l[i], array_l[i + 1]) for i in range(len(array_l))
if i % 2 == 0]
if __name__ == '__main__':
#import doctest
#doctest.testmod()
from kyotocabinet import DB
db = DB()
db.open('/mnt/zdata/kyoto/kyoto.kch', DB.OREADER)
print db.get(15439)
class DataStorage(object):
"""
Parent class for RowData and KeyValueData.
"""
def __init__(self, filename, headers=None):
self.filename = filename
self.ext = os.path.splitext(filename)[1]
self.headers = headers
if os.path.exists(self.filename):
self.init_read()
else:
self.init_write()
def init_write(self):
self.mode = "write"
if self.ext == ".csv":
self._data_file = open(self.filename, "wb")
self._writer = csv.writer(self._data_file)
if self.headers:
self._writer.writerow(self.headers)
elif self.ext == ".json":
self._storage = {}
elif self.ext == ".kch":
from kyotocabinet import DB
self._storage = DB()
if not self._storage.open(self.filename, DB.OWRITER | DB.OCREATE):
msg = "Error opening kyotocabinet db: %s" % (self._storage.error())
raise dexy.commands.UserFeedback(msg)
elif self.ext == ".sqlite3":
self.init_write_sqlite3()
else:
raise dexy.commands.UserFeedback("unsupported extension %s" % self.ext)
def init_read(self):
self.mode = "read"
if self.ext == ".csv":
self._file = open(self.filename, "rb")
elif self.ext == ".json":
with open(self.filename, "rb") as f:
self._storage = json.load(f)
elif self.ext == ".kch":
from kyotocabinet import DB
self._storage = DB()
self._storage.open(self.filename, DB.OREADER)
elif self.ext == ".sqlite3":
import sqlite3
self._storage = sqlite3.connect(self.filename)
self._cursor = self._storage.cursor()
else:
raise dexy.commands.UserFeedback("unsupported extension %s" % self.ext)
def save(self):
if self.ext == ".csv":
self._data_file.close()
elif self.ext == ".json":
with open(self.filename, "wb") as f:
import json
json.dump(self._storage, f)
elif self.ext == ".kch":
if not self._storage.close():
raise dexy.commands.UserFeedback(self._storage.error())
elif self.ext == ".sqlite3":
self._storage.commit()
self._cursor.close()
else:
raise dexy.commands.UserFeedback("unsupported extension %s" % self.ext)
def get_user_stats(username, db_file):
item = {
"username": username,
"aliases": [],
"total_posts": 0,
"domains": {},
"first_post_date": None,
"first_post_date_unix": None,
"most_recent_post": None,
"most_recent_post_unix": 0,
"average_posts_per_hour": 0.0,
"average_posts_per_day": 0.0,
"average_posts_per_week": 0.0
}
db = DB()
if not db.open("{0}".format(db_file), DB.OREADER | DB.OCREATE):
print "Could not open database."
cur = db.cursor()
cur.jump()
while True:
rec = cur.get(False)
if not rec:
break
loaded_rec = loads(rec[1])
if loaded_rec['person'] != username:
cur.step()
continue
# Looks like this is a post by the user we're looking for
split = get_domain(loaded_rec)
if item['domains'].get(split, False) == False:
item['domains'][split] = 1
else:
item['domains'][split] = item['domains'][split] + 1
if item['first_post_date_unix'] is None:
item['first_post_date_unix'] = loaded_rec['created_at']
if item['most_recent_post_unix'] < loaded_rec['created_at']:
item['most_recent_post_unix'] = loaded_rec['created_at']
item['total_posts'] = item['total_posts'] + 1
cur.step()
cur.disable()
db.close()
# Clean up everything
first_time = None
if item['first_post_date_unix'] is not None:
unix = float(item['first_post_date_unix'])
first_time = datetime.fromtimestamp(unix)
item['first_post_date'] = first_time.isoformat()
recent_time = None
if item['most_recent_post_unix'] is not None:
unix = float(item['most_recent_post_unix'])
recent_time = datetime.fromtimestamp(unix)
item['most_recent_post'] = recent_time.isoformat()
if first_time and recent_time:
delta = recent_time - first_time
item['user_age_days'] = delta.days
item['user_age_seconds'] = delta.total_seconds()
item['average_posts_per_hour'] = item['total_posts'] / (
delta.total_seconds() / 60.0)
item['average_posts_per_day'] = item['total_posts'] / (
delta.total_seconds() / 60.0 / 24.0)
item['average_posts_per_week'] = item['total_posts'] / (
delta.total_seconds() / 60.0 / 24.0 / 7.0)
return item
return izip(a, b)
#for k,v in pairwise(range(10)):
# print k,v
def convert2array(dict_value):
'''
>>> convert2array({1:0.1,2:0.3})
array('L', [1L, 429496729L, 2L, 1288490188L])
'''
result_list = []
for k,v in dict_value:
result_list.extend([k,int(v*MAX_INT)])
result = array('L',result_list)
return result
def convert2dict(array_l):
'''
>>> convert2dict([1L, 429496729L, 2L, 1288490188L])
{1:0.1,2:0.3}
'''
return [ (array_l[i],array_l[i+1]) for i in range(len(array_l)) if i%2==0]
if __name__=='__main__':
#import doctest
#doctest.testmod()
from kyotocabinet import DB
db = DB()
db.open('/mnt/zdata/kyoto/kyoto.kch', DB.OREADER)
print db.get(15439)
class KyotoIndex(BinaryIndex):
''' Kyoto Cabinet index.
Notably this uses a B+ tree for the index and thus one can
traverse from one key forwards and backwards, which supports
the coming Store synchronisation processes.
'''
NAME = 'kyoto'
SUFFIX = 'kct'
def __init__(self, nmdbpathbase):
super().__init__(nmdbpathbase)
self._kyoto = None
@classmethod
def is_supported(cls):
''' Test whether this index class is supported by the Python environment.
'''
# pylint: disable=import-error,unused-import,import-outside-toplevel
try:
import kyotocabinet
except ImportError:
return False
return True
def startup(self):
''' Open the index.
'''
# pylint: disable=import-error,import-outside-toplevel
from kyotocabinet import DB
self._kyoto = DB()
self._kyoto.open(self.path, DB.OWRITER | DB.OCREATE)
def shutdown(self):
''' Close the index.
'''
self._kyoto.close()
self._kyoto = None
def flush(self):
''' Flush pending updates to the index.
'''
try:
self._kyoto.synchronize(hard=False)
except TypeError:
self._kyoto.synchronize()
def __len__(self):
return self._kyoto.count()
def __contains__(self, key):
return self._kyoto.check(key) >= 0
def __getitem__(self, key):
binary_entry = self._kyoto.get(key)
if binary_entry is None:
raise KeyError(key)
return binary_entry
def __setitem__(self, key, binary_entry):
self._kyoto[key] = binary_entry
def keys(self, *, start_hashcode=None):
''' Generator yielding the keys from the index
in order starting with optional `start_hashcode`.
Parameters:
* `start_hashcode`: the starting key; if missing or `None`,
iteration starts with the first key in the index
'''
cursor = self._kyoto.cursor()
if start_hashcode is not None:
cursor.jump(start_hashcode)
yield cursor.get_key()
while cursor.step():
yield cursor.get_key()
cursor.disable()
sorted_keys = keys
__iter__ = keys
class FeatureSelector( Frontend ):
def __init__( self, fn, mode ):
Frontend.__init__( self, fn, mode );
self._kdbfn = None;
self._kdb = None;
self._ldbdn = None;
self._ldb = None;
self._len_c = None;
self._len_b = None;
self._len_x = None;
self._ic = None;
self._icbp = None;
self._needs_initialization = True;
self._core_dims = set();
self._satellite_dims = set();
self._removed_dims = set();
self._remove_c = set();
self._remove_b = set();
self._remove_x = set();
self.bypass_c = False;
self.bypass_b = False;
self.bypass_x = False;
def __enter__( self ):
if self._mode == "r":
with open( self._fn, "rb" ) as f:
state = pickle_load( f );
self._len_c = state[ "c" ];
self._len_b = state[ "b" ];
self._len_x = state[ "x" ];
self._lenrow = self._len_c + self._len_b + self._len_x;
self._ic = state[ "ic" ];
self._icbp = state[ "icbp" ];
if self._mode == "w":
with NamedTemporaryFile() as tmpfn:
self._kdbfn = tmpfn.name + '.kch';
self._kdb = KDB();
try:
assert self._kdb.open( self._kdbfn, KDB.OWRITER | KDB.OCREATE );
except:
print( str( self._kdb.error() ) );
raise;
with TemporaryDirectory() as tmpdirname:
self._ldbdn = tmpdirname;
self._ldb = LDB( self._ldbdn, create_if_missing=True );
return self;
def __exit__( self, exc_type, exc_value, traceback ):
assert Frontend.__exit__( self, exc_type, exc_value, traceback ) == False;
if self._ldb is not None:
sleep( 3.0 );
self._ldb.close()
if self._ldbdn is not None:
rmtree( self._ldbdn );
if self._kdb is not None:
try:
assert self._kdb.close();
except:
print( str( self._kdb.error() ) );
raise;
if self._kdbfn is not None:
remove( self._kdbfn );
def train( self, row ):
( y, c, b, x ) = row;
if self._len_c is None:
self._len_c = len(c);
assert self._len_c == len(c);
if self._len_b is None:
self._len_b = len(b);
assert self._len_b == len(b);
if self._len_x is None:
self._len_x = len(x);
assert self._len_x == len(x);
row = c + b + x;
if Frontend.train( self, row ):
return True;
keyfmt = '>IIIII';
for i in range( 0, self._lenrow ):
for j in range( 0, self._lenrow ):
if ( i >= j ) and ( not ( i == self._lenrow-1 ) ):
continue;
key = pack( keyfmt, i, j, y, row[i], row[j] );
try:
assert self._kdb.increment( key, 1, 0 );
except:
print( str(self._kdb.error()) );
raise;
def _stats( self, cnt_by_a, cnt_by_b, cnt_by_ab ):
h_a = 0.0;
h_b = 0.0;
h_ab = 0.0;
for ( val_a, cnt ) in cnt_by_a.items():
p = float(cnt) / float(self._rowcount);
if p > 0.0:
h_a -= p * log( p, 2.0 );
for ( val_b, cnt ) in cnt_by_b.items():
p = float(cnt) / float(self._rowcount);
if p > 0.0:
h_b -= p * log( p, 2.0 );
for( (val_a,val_b), cnt ) in cnt_by_ab.items():
p = float(cnt) / float(self._rowcount);
if p > 0.0:
h_ab -= p * log( p, 2.0 );
if h_a == 0.0:
return 1.0;
if h_b == 0.0:
return 1.0;
mi = h_a + h_b - h_ab;
return ( mi / min( h_a, h_b ), h_a, h_b, h_ab, mi );
def _get_info_content_by_dimension( self, i ):
keyfmt = '>IIIII';
valfmt = '>Q';
j = None;
cnt_by_a = {};
cnt_by_b = {};
cnt_by_ab = {};
total = 0;
with self._ldb.iterator() as it:
it.seek( pack( keyfmt, i,0,0,0,0 ) );
for ( key, val ) in it:
key = unpack( keyfmt, key );
val = unpack( valfmt, val )[ 0 ];
if not ( key[0] == i ):
break;
if j is None:
j = key[1];
if not ( key[1] == j ):
break;
# key[2] is the y-value
a = key[2];
# key[3] is the value for the i-th dimension
b = key[3];
cnt_by_ab[ (a,b) ] = cnt_by_ab.get( (a,b), 0 ) + val;
cnt_by_a[ a ] = cnt_by_a.get( a, 0 ) + val;
cnt_by_b[ b ] = cnt_by_b.get( b, 0 ) + val;
total += val;
try:
assert total == self._rowcount;
except:
print( i, j, total, self._rowcount );
raise;
return self._stats( cnt_by_a, cnt_by_b, cnt_by_ab );
def _get_info_content_by_pair( self, i, j ):
keyfmt = '>IIIII';
valfmt = '>Q';
cnt_by_a = {};
cnt_by_b = {};
cnt_by_ab = {};
total = 0;
with self._ldb.iterator() as it:
it.seek( pack( keyfmt, i,j,0,0,0 ) );
for ( key, val ) in it:
key = unpack( keyfmt, key );
val = unpack( valfmt, val )[ 0 ];
if not ( ( key[0] == i ) and ( key[1] == j ) ):
break;
# key[2] is the y-value, key[3] the i-th value for the i-th dim
a = ( key[2], key[3] );
# key[2] is the y-value, key[4] the i-th value for the j-th dim
b = ( key[2], key[4] );
assert (a,b) not in cnt_by_ab;
cnt_by_ab[ (a,b) ] = cnt_by_ab.get( (a,b), 0 ) + val;
cnt_by_a[ a ] = cnt_by_a.get( a, 0 ) + val;
cnt_by_b[ b ] = cnt_by_b.get( b, 0 ) + val;
total += val;
assert total == self._rowcount;
return self._stats( cnt_by_a, cnt_by_b, cnt_by_ab );
def _finalize( self ):
assert Frontend._finalize( self ) is None;
if False:
print( "unique combinations = ", self._kdb.count() );
keyfmt = '>IIIII';
valfmt = '>Q';
c = self._kdb.cursor();
c.jump();
gt2 = 0;
gt4 = 0;
gt8 = 0;
gt16 = 0;
gt32 = 0;
while True:
r = c.get( True );
if not r:
break;
self._ldb.put( r[0], r[1] );
key = unpack( keyfmt, r[0] );
val = unpack( valfmt, r[1] )[ 0 ];
if val > 2:
gt2 += 1;
if val > 4:
gt4 += 1;
if val > 8:
gt8 += 1;
if val > 16:
gt16 += 1;
if val > 32:
gt32 += 1;
if False:
print( gt2, gt4, gt8, gt16, gt32 );
self._ic = {};
for i in range( 0, self._lenrow ):
self._ic[ i ] = self._get_info_content_by_dimension( i );
self._icbp = {};
for i in range( 0, self._lenrow ):
for j in range( 0, self._lenrow ):
if i >= j:
continue;
self._icbp[ (i,j) ] = self._get_info_content_by_pair( i, j );
self._state \
= { "ic": self._ic,
"icbp": self._icbp,
"c": self._len_c,
"b": self._len_b,
"x": self._len_x };
def _fmt_dim( self, d_ ):
d = None;
if d_ < self._len_c:
d = "c" + str( d_ );
elif d_ < self._len_c + self._len_b:
d = "b" + str( d_ - self._len_c );
elif d_ < self._len_c + self._len_b + self._len_x:
d = "x" + str( d_ - self._len_c - self._len_b );
else:
assert False;
return "{:d}({:s})".format( d_, d );
def _init( self ):
self._needs_initialization = False;
if False:
for i in sorted( self._ic ):
(corr,h_a,h_b,h_ab,mi) = self._ic[ i ];
print(
"{:s} {:1.4f} {:1.4f} {:1.4f} {:1.4f} {:1.4f}"\
.format(
self._fmt_dim( i ),
corr,
h_a,
h_b,
h_ab,
mi
)
);
for (i,j) in sorted( self._icbp ):
(corr,h_a,h_b,h_ab,mi) = self._icbp[ (i,j) ];
print(
"{:s} {:s} {:1.4f} {:1.4f} {:1.4f} {:1.4f} {:1.4f}"\
.format(
self._fmt_dim( i ),
self._fmt_dim( j ),
corr,
h_a,
h_b,
h_ab,
mi
)
);
entropy \
= [ ( h_ab, i ) \
for ( i, (corr,h_a,h_b,h_ab,mi) ) in self._ic.items() ];
output_correlation \
= [ ( corr, i ) \
for ( i, (corr,h_a,h_b,h_ab,mi) ) in self._ic.items() ];
self._core_dims = set();
self._core_dims \
|= { i \
for ( h_ab, i ) \
in sorted( entropy, reverse=True )[ :5 ] };
self._core_dims \
|= { i \
for ( h_ab, i ) \
in sorted( output_correlation, reverse=True )[ :3 ] };
if True:
print(
"core = ",
" ".join([ self._fmt_dim(d) for d in self._core_dims ])
);
self._satellite_dims = set();
for core_dim in self._core_dims:
satellite_dim = None;
satellite_dim_c = None;
satellite_dim_stats = None;
for ( (i,j), (corr,h_a,h_b,h_ab,mi) ) in self._icbp.items():
if corr <= 0.5:
continue;
other_dim = None;
if i == core_dim:
other_dim = j;
elif j == core_dim:
other_dim = i;
else:
continue;
if ( satellite_dim_c is None ) or ( corr > satellite_dim_c ):
satellite_dim = other_dim;
satellite_dim_c = corr;
satellite_dim_stats = (corr,h_a,h_b,h_ab,mi);
if satellite_dim is not None:
self._satellite_dims.add( satellite_dim );
if False:
print(
'->',
self._fmt_dim(core_dim),
self._fmt_dim(satellite_dim)
);
print(
"{:1.4f} {:1.4f} {:1.4f} {:1.4f} {:1.4f}"\
.format( *(corr,h_a,h_b,h_ab,mi) )
);
if True:
print(
"satellite = ",
" ".join([ self._fmt_dim(d) for d in self._satellite_dims ])
);
self._removed_dims = set();
for i in self._ic:
if i not in self._core_dims and i not in self._satellite_dims:
self._removed_dims.add( i );
if True:
print(
"removed = ",
" ".join([ self._fmt_dim(d) for d in self._removed_dims ])
);
for d_ in self._removed_dims:
if d_ < self._len_c:
self._remove_c.add( d_ );
elif d_ < self._len_c + self._len_b:
self._remove_b.add( d_ - self._len_c );
elif d_ < self._len_c + self._len_b + self._len_x:
self._remove_x.add( d_ - self._len_c - self._len_b );
else:
assert False;
def apply_c( self, c ):
if self.bypass_c:
return c;
if self._needs_initialization:
self._init();
c_ = [];
for ( i, cval ) in enumerate( c ):
if not i in self._remove_c:
c_.append( cval );
return c_;
def apply_b( self, b ):
if self.bypass_b:
return b;
if self._needs_initialization:
self._init();
b_ = [];
for ( i, bval ) in enumerate( b ):
if not i in self._remove_b:
b_.append( bval );
return b_;
def apply_x( self, x ):
if self.bypass_x:
return x;
if self._needs_initialization:
self._init();
x_ = [];
for ( i, xval ) in enumerate( x ):
if not i in self._remove_x:
x_.append( xval );
return x_;
def __call__( self, row ):
if self._needs_initialization:
self._init();
( y, c, b, x ) = row;
y_ = y;
return \
( y_,
self.apply_c( c ),
self.apply_b( b ),
self.apply_x( x ) );
from kyotocabinet import DB
from collections import defaultdict
from array import array
from zkit.zitertools import chunkiter
from operator import itemgetter
from zdata.tag.name2id import NAME2ID
from zkit.txt_cleanup import sp_txt
ID2NAME = defaultdict(list)
for name, id in NAME2ID.iteritems():
ID2NAME[id].append(name)
db_tag_bayes = DB()
db_tag_bayes.open(join(ZDATA_PATH, "data/bayes.kch"), DB.OREADER)
def tag_id_rank_list_by_txt(txt):
txt = txt.lower()
tag_id_list_rank = defaultdict(int)
for word, rank in tf_idf_seg_txt(txt):
#print word
ars = db_tag_bayes.get(word)
if ars:
ar = array('I')
ar.fromstring(ars)
#print len(ar)
#print db_tag_bayes[word]
#print word, ar
for tag_id, bayes in chunkiter(ar, 2):
class KyotoCabinetGraph(BaseGraph):
def __init__(self, path):
# create the database object
self._path = path
self._db = DB()
# open the database
if not self._db.open(path, DB.OREADER | DB.OWRITER | DB.OCREATE):
raise GrapheekDataKyotoCabinetInitFailureException(
str(self._db.error()))
super(KyotoCabinetGraph, self).__init__()
self._ensure_prepared()
self._closed = False
# Start method overriding :
def _db_close(self):
if not self._closed:
self._db.close()
def _transaction_begin(self):
self._db.begin_transaction()
return True
def _transaction_commit(self, txn):
self._db.end_transaction(True)
def _transaction_rollback(self, txn):
self._db.end_transaction(False)
def _has_key(self, key):
return self._db.check(key) >= 0
def _get(self, txn, key):
raw_data = self._db.get(key)
if raw_data is None:
return UNDEFINED # Not returning None, as None is a valid value
return msgpack.loads(raw_data, encoding='utf8')
def _bulk_get(self, txn, keys):
result = {}
key_raw_datas = self._db.get_bulk(keys)
for key, raw_data in list(key_raw_datas.items()):
if PYTHON2: # pragma : no cover
k = key
else: # pragma : no cover
k = str(key, encoding='utf8')
result[k] = msgpack.loads(raw_data, encoding='utf8')
return result
def _set(self, txn, key, value):
res = self._db.set(key, msgpack.dumps(value, encoding='utf8'))
if not (res): # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
def _bulk_set(self, txn, updates):
dic = {}
for key, value in list(updates.items()):
dic[key] = msgpack.dumps(value, encoding='utf8')
res = self._db.set_bulk(dic)
if res == -1: # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
def _remove(self, txn, key):
# Contrary to LocalMemoryGraph implementation, it is not needed to wrap
# key removal in try.. except because KyotoCabinet only send "False"
# when key does not exist
# Thus ... _removemethod is idempotent (cf LocalMemoryGraph _remove method comment)
self._db.remove(key)
def _bulk_remove(self, txn, keys):
res = self._db.remove_bulk(list(keys))
if res == -1: # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
def _remove_prefix(self, txn, prefix):
keys = self._db.match_prefix(prefix)
self._db.remove_bulk(keys)
# overriding list methods
# looks like a bucket of hacks, and yes indeed it is :)
# btw, it REALLY improves performance if we compare to default implementation which,
# in the case of KyotoCabinet would involve msgpack deserialization followed by a serialization
def _init_lst(self, txn, key):
res = self._db.set(key, '')
if not (res): # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
def _get_lst(self, txn, key):
value = self._db.get(key)
if value is None:
return UNDEFINED
# look _append_to_lst code below to understand why a split is done
# And why resulting list is sliced from 1
if PYTHON2: # pragma : no cover
return list(map(int, value.split('|')[1:]))
return list(map(
int,
str(value, encoding='utf8').split('|')[1:])) # pragma : no cover
def _set_lst(self, txn, key, values):
newval = '|'.join([str(value) for value in values])
res = self._db.set(key, '|' + newval)
if not (res): # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
def _bulk_get_lst(self, txn, keys):
dic_values = self._db.get_bulk(keys)
results = []
for key in keys:
if PYTHON2: # pragma : no cover
values = dic_values.get(key, UNDEFINED)
else: # pragma : no cover
values = dic_values.get(bytes(key, encoding='utf8'), UNDEFINED)
if values == UNDEFINED:
results.append([])
else:
if PYTHON2: # pragma : no cover
results.append(list(map(int, values.split('|')[1:])))
else: # pragma : no cover
results.append(
list(
map(int,
str(values, encoding='utf8').split('|')[1:])))
return results
def _append_to_lst(self, txn, key, value):
self._db.append(key, '|' + str(value))
def _bulk_append_to_lst(self, txn, key, values):
newval = '|'.join([str(value) for value in values])
self._db.append(key, '|' + newval)
def _remove_from_lst(self, txn, key, value):
old = self._db.get(key)
if not PYTHON2: # pragma : no cover
old = str(old, encoding='utf8')
# Caution : we are only removing ONE occurence
# This is voluntary
# For instance, it lst contains neighbour node, we need to remove only one occurence
# cause current entity and neighbour node can be linked multiple time
new = old.replace('|%s' % value, '', 1)
if new == old:
raise ValueError("list.remove(x): x not in list")
res = self._db.set(key, new)
if not (res): # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
def _bulk_remove_from_lst(self, txn, key, values):
assert (len(values))
old = self._db.get(key)
if PYTHON2: # pragma : no cover
new = old
else: # pragma : no cover
new = str(old, encoding='utf8')
for value in values:
new = new.replace('|%s' % value, '', 1)
if new == old: # pragma : no cover
raise ValueError("list.remove(x): x not in list")
res = self._db.set(key, new)
if not (res): # pragma : no cover
raise GrapheekDataKyotoCabinetException(
'KyotoCabinet : error while saving')
return res
class BKNNModel( Model ):
def __init__( self, fn, mode, catfe, binfe, contfe, fdisc, fsel, kval ):
Model.__init__( self, fn, mode, catfe, binfe, contfe, fdisc, fsel );
self._kval = kval;
self._fn_cdata = self._fn;
self._fn_ddata = self._fn.replace( '.kch', '-discrete.kch' );
self._fn_meta = self._fn.replace( '.kch', '-meta.pickle' );
self._fn_icov = self._fn.replace( '.kch', '-icov.pickle' );
self._cdata = None;
self._ddata = None;
self._len_c = None;
self._len_b = None;
self._len_x = None;
self._rowcount = None;
self._total_pos = None;
self._total_neg = None;
self._icov = None;
self._co = None;
self._sample_y = [];
self._sample_c = [];
self._sample_b = [];
self._sample_x = [];
self._sample_x_ = [];
self._needs_finalization = False;
self._needs_initialization = True;
self._dmarginals = {};
self._dscores = {};
self._sparse_points = 0;
self._bias = None;
def __enter__( self ):
self._cdata = DB();
self._ddata = DB();
try:
if self._mode == "r":
assert self._cdata.open( self._fn_cdata, DB.OREADER );
elif self._mode == "w":
if isfile( self._fn_cdata ):
remove( self._fn_cdata );
assert self._cdata.open( self._fn_cdata, DB.OWRITER | DB.OCREATE );
else:
assert False;
except:
if self._cdata is not None:
print( str( self._cdata.error() ) );
raise;
try:
if self._mode == "r":
assert self._ddata.open( self._fn_ddata, DB.OREADER );
elif self._mode == "w":
if isfile( self._fn_ddata ):
remove( self._fn_ddata );
assert self._ddata.open( self._fn_ddata, DB.OWRITER | DB.OCREATE );
else:
assert False;
except:
if self._ddata is not None:
print( str( self._ddata.error() ) );
raise;
if self._mode == "r":
with open( self._fn_meta, 'rb' ) as f:
r = pickle_load( f );
self._len_c = r[ "c" ];
self._len_b = r[ "b" ];
self._len_x = r[ "x" ];
self._co = r[ "co" ];
with open( self._fn_icov, 'rb' ) as f:
self._icov = pickle_load( f );
return self;
def __exit__( self, exc_type, exc_value, traceback ):
ex_w_exc = False;
ex_w_exc = ex_w_exc or ( exc_type is not None );
ex_w_exc = ex_w_exc or ( exc_value is not None );
ex_w_exc = ex_w_exc or ( traceback is not None );
if ( not ex_w_exc ) and ( self._mode == "w" ):
if self._needs_finalization:
self._finalize();
with open( self._fn_meta, 'wb' ) as f:
r = { "c": self._len_c,
"b": self._len_b,
"x": self._len_x,
"co": self._co };
pickle_dump( r, f );
with open( self._fn_icov, 'wb' ) as f:
pickle_dump( self._icov, f );
if self._cdata is not None:
try:
assert self._cdata.close();
except:
print( str( self._cdata.error() ) );
raise;
self._cdata = None;
if self._ddata is not None:
try:
assert self._ddata.close();
except:
print( str( self._ddata.error() ) );
raise;
self._ddata = None;
if ex_w_exc and ( self._mode == "w" ):
if isfile( self._fn_cdata ):
remove( self._fn_cdata );
if isfile( self._fn_ddata ):
remove( self._fn_ddata );
if isfile( self._fn_meta ):
remove( self._fn_meta );
if isfile( self._fn_icov ):
remove( self._fn_icov );
return False;
def train( self, row ):
self._needs_finalization = True;
( y, c, b, x ) = row;
c = self._fsel.apply_c( self._catfe( c ) );
b = self._fsel.apply_b( self._binfe( b ) );
x = self._contfe( x );
x_ = self._fdisc( x );
x = self._fsel.apply_x( x );
x_ = self._fsel.apply_x( x_ );
if False:
print( y, c, b, x, x_ );
if self._len_c is None:
self._len_c = len(c);
assert self._len_c == len(c);
if self._len_b is None:
self._len_b = len(b);
assert self._len_b == len(b);
if self._len_x is None:
self._len_x = len(x);
assert self._len_x == len(x);
if self._rowcount is None:
self._rowcount = 0;
self._rowcount += 1;
dkeyfmt = '>' + ( 'I' * ( 1 + self._len_c + self._len_b ) );
self._ddata.increment( pack( dkeyfmt, y, *(c+b) ), 1, 0 );
ckeyfmt = '>' + ( 'I' * len(x) );
cvalfmt = '>I' + ( 'f' * len(x) );
self._cdata.append( pack( ckeyfmt, *x_ ), pack( cvalfmt, y, *x ) );
if len( self._sample_x ) < 50000:
assert len( self._sample_x ) == len( self._sample_y );
assert len( self._sample_x ) == len( self._sample_c );
assert len( self._sample_x ) == len( self._sample_b );
assert len( self._sample_x ) == len( self._sample_x_ );
self._sample_y.append( y );
self._sample_c.append( c );
self._sample_b.append( b );
self._sample_x.append( x );
self._sample_x_.append( x_ );
return False;
def _init( self ):
self._needs_initialization = False;
c = self._ddata.cursor();
c.jump();
keyfmt = '>' + ( 'I' * ( 1 + self._len_c + self._len_b ) );
valfmt = '>Q';
while True:
r = c.get( True );
if not r:
break;
dbkey = unpack( keyfmt, r[0] );
dbval = unpack( valfmt, r[1] )[ 0 ];
additional_count = dbval;
y = dbkey[ 0 ];
for ( i, value_of_variable_i ) in enumerate( dbkey[ 1: ] ):
if not i in self._dmarginals:
self._dmarginals[ i ] = {};
self._dmarginals[ i ][ (y,value_of_variable_i) ] \
= self._dmarginals[ i ].get( (y,value_of_variable_i), 0 ) \
+ additional_count;
for ( i, count_by_val ) in self._dmarginals.items():
total = 0;
total_neg = 0;
total_pos = 0;
for ( ( y, val ), cnt ) in count_by_val.items():
total += cnt;
if y == 0:
total_neg += cnt;
elif y == 1:
total_pos += cnt;
if self._rowcount is None:
self._rowcount = total;
assert self._rowcount == total;
if self._total_neg is None:
self._total_neg = total_neg;
try:
assert self._total_neg == total_neg;
except:
print( self._total_neg, total_neg );
raise;
if self._total_pos is None:
self._total_pos = total_pos;
try:
assert self._total_pos == total_pos;
except:
print( self._total_pos, total_pos );
raise;
assert ( self._total_pos + self._total_neg ) == self._rowcount;
for i in self._dmarginals:
values = set([ val for (y,val) in self._dmarginals[ i ].keys() ]);
if i not in self._dscores:
self._dscores[ i ] = {};
for val in values:
pos_cnt = self._dmarginals[ i ].get( (1,val), 0 );
neg_cnt = self._dmarginals[ i ].get( (0,val), 0 );
p_pos \
= log( float(pos_cnt) + SMOOTHING, 2.0 ) \
- log( float(self._total_pos) + float( len(values) ) * SMOOTHING, 2.0 );
p_neg \
= log( float(neg_cnt) + SMOOTHING, 2.0 ) \
- log( float(self._total_neg) + float( len(values) ) * SMOOTHING, 2.0 );
self._dscores[ i ][ val ] = p_pos - p_neg;
p_pos \
= log( float(self._total_pos), 2.0 ) \
- log( float(self._rowcount), 2.0 );
p_neg \
= log( float(self._total_neg), 2.0 ) \
- log( float(self._rowcount), 2.0 );
self._bias = p_pos - p_neg;
if False:
for i in sorted( self._dscores.keys() ):
score_by_val = self._dscores[ i ];
for ( val, score ) in score_by_val.items():
print( "{:d} {:10d} {:+2.4f}".format( i, val, score ) );
def _apply( self, row ):
if self._needs_initialization:
self._init();
( c, b, x, x_ ) = row;
ckeyfmt = '>' + ( 'I' * len(x_) );
cvalfmt = '>I' + ( 'f' * len(x) );
cvalsz = calcsize( cvalfmt );
rng = [];
for xval in x_:
rng.append(
[ xv \
for xv \
in [ xval-2, xval-1, xval, xval+1, xval+2 ] \
if 0 <= xv <= 31 ]
);
x_vec = np.array( x ).reshape( 1, self._len_x ).T;
nearest_positive = [];
all_negative = [];
found_ident = 0;
for xvals in product( *rng ):
try:
ckey = pack( ckeyfmt, *xvals );
except:
print( ckeyfmt, xvals );
raise;
val = self._cdata.get( ckey );
while val:
if len(val) <= cvalsz:
assert len(val) == cvalsz;
val_ = val[:cvalsz];
val = val[cvalsz:];
pt = unpack( cvalfmt, val_ );
pt_y = pt[0];
pt_x = pt[1:];
pt_x_vec = np.array( pt_x ).reshape( 1, self._len_x ).T;
diff = pt_x_vec - x_vec;
dist = np.sqrt( np.dot( np.dot( diff.T, self._icov ), diff ) );
if dist <= 0.0001:
found_ident += 1;
continue;
if pt_y == 0:
all_negative.append( dist );
continue;
assert pt_y == 1;
nearest_positive.append( dist );
nearest_positive.sort();
nearest_positive = nearest_positive[:self._kval];
# assert found_ident == 1;
# assert len( nearest_positive ) == self._kval;
if len( nearest_positive ) < self._kval:
self._sparse_points += 1;
score = self._bias;
# if len( nearest_positive ) > 0:
if True:
if len( nearest_positive ) == 0:
threshold = None;
else:
threshold = nearest_positive[-1];
neg_cnt = 0;
for dist in all_negative:
if ( threshold is None ) or ( dist <= threshold ):
neg_cnt += 1;
p_pos \
= log( float( len(nearest_positive) ) + SMOOTHING, 2.0 ) \
- log( float(self._total_pos) + 2.0 * SMOOTHING, 2.0 );
p_neg \
= log( float(neg_cnt) + SMOOTHING, 2.0 ) \
- log( float(self._total_neg) + 2.0 * SMOOTHING, 2.0 );
score += p_pos - p_neg;
for ( i, dval ) in enumerate( c+b ):
score += self._dscores[ i ].get( dval, 0.0 );
if self._co is None:
return score;
else:
if score >= self._co:
return 1;
else:
return 0;
def _finalize( self ):
self._needs_finalization = False;
covsample = np.array( self._sample_x );
cov = np.cov( covsample.T );
self._icov = LA.inv( cov );
sample \
= zip(
self._sample_c,
self._sample_b,
self._sample_x,
self._sample_x_
);
scores = [];
for ( c, b, x, x_ ) in sample:
scores.append( self._apply( [ c, b, x, x_ ] ) );
sorted_scores = list( sorted( scores ) );
cutoffs = [];
for idx in range(0,1000):
ratio = float(idx) / 1000.0;
cutoffs.append(
sorted_scores[ int( float( len(sorted_scores) ) * ratio ) ]
);
if False:
pprint( cutoffs );
stats_by_co = [];
for coidx in range( 0, len(cutoffs) ):
stats_by_co.append( { "tp": 0, "fp": 0, "tn": 0, "fn": 0 } );
for ( y, score ) in zip( self._sample_y, scores ):
for ( coidx, co ) in enumerate( cutoffs ):
if score >= co:
if y == 1:
stats_by_co[ coidx ][ "tp" ] += 1;
else:
assert y == 0;
stats_by_co[ coidx ][ "fp" ] += 1;
else:
if y == 0:
stats_by_co[ coidx ][ "tn" ] += 1;
else:
assert y == 1;
stats_by_co[ coidx ][ "fn" ] += 1;
max_fscore = None;
max_fscore_coidx = None;
for ( coidx, co ) in enumerate( cutoffs ):
tp = stats_by_co[ coidx ][ "tp" ];
fp = stats_by_co[ coidx ][ "fp" ];
tn = stats_by_co[ coidx ][ "tn" ];
fn = stats_by_co[ coidx ][ "fn" ];
if (tp+fp) <= 0:
continue;
if (tp+fn) <= 0:
continue;
precision = float(tp) / float(tp+fp);
recall = float(tp) / float(tp+fn);
if (precision+recall) <= 0.0:
continue;
fscore = 2.0 * ( ( precision * recall ) / ( precision + recall ) );
if ( max_fscore is None ) or ( fscore > max_fscore ):
max_fscore = fscore;
max_fscore_coidx = coidx;
assert max_fscore_coidx is not None;
self._co = cutoffs[ max_fscore_coidx ];
# assert self._sparse_points == 0;
if True:
print( self._sparse_points );
print( self._co );
print( max_fscore );
def __call__( self, row ):
( c, b, x ) = row;
c = self._fsel.apply_c( self._catfe( c ) );
b = self._fsel.apply_b( self._binfe( b ) );
x = self._contfe( x );
x_ = self._fdisc( x );
x = self._fsel.apply_x( x );
x_ = self._fsel.apply_x( x_ );
try:
assert self._len_c == len(c);
assert self._len_b == len(b);
assert self._len_x == len(x);
assert self._len_x == len(x_);
except:
print( self._len_c, self._len_b, self._len_x );
raise;
return self._apply( ( c, b, x, x_ ) );