def token_series(cls, token, corpus=None, pos=None):
"""Get an offset -> count series for a word.
Args:
token (str)
corpus (str)
pos (str)
Returns: OrderedDict
"""
query = (session.query(cls.bin, func.sum(cls.count)).filter(
cls.token == token).group_by(cls.bin).order_by(cls.bin))
if corpus:
query = query.filter(cls.corpus == corpus)
if pos:
query = query.filter(cls.pos == pos)
series = np.zeros(100)
for offset, count in query:
series[offset] = count
return series
def pos_tags(cls):
"""Get a list of all POS tags.
Returns: set
"""
query = session.query(distinct(cls.pos))
return sorted([r[0] for r in query.all()])
def token_counts(cls, min_count=0):
"""Get total (un-bucketed) token counts.
Args:
min_count (int)
Returns: OrderedDict
"""
query = (session.query(cls.token, func.sum(cls.count)).group_by(
cls.token).having(func.sum(cls.count) > min_count).order_by(
func.sum(cls.count).desc()))
return OrderedDict(query.all())
def token_pos_counts(cls, min_count=0):
"""Get total (token, pos) counts.
Args:
min_count (int)
Returns: OrderedDict
"""
query = (session.query(
cls.token, cls.pos, func.sum(cls.count)).group_by(
cls.token,
cls.pos).having(func.sum(cls.count) > min_count).order_by(
func.sum(cls.count).desc()))
return OrderedDict([((token, pos), count)
for token, pos, count in query.all()])
def pos_series(cls, tag, corpus=None):
"""Get an offset -> count series for a POS tag.
Args:
tag (str)
corpus (str)
Returns: OrderedDict
"""
query = (session.query(cls.bin, func.sum(cls.count)).filter(
cls.pos == tag).group_by(cls.bin).order_by(cls.bin))
if corpus:
query = query.filter(cls.corpus == corpus)
series = np.zeros(100)
for offset, count in query:
series[offset] = count
return series