spaCy/spacy/gold.pyx

200 lines
6.3 KiB
Cython
Raw Normal View History

2015-03-09 08:46:22 +03:00
import numpy
import codecs
2015-05-06 17:27:31 +03:00
import json
import random
import re
2015-05-27 04:17:50 +03:00
from spacy.munge.read_ner import tags_to_entities
2015-03-09 08:46:22 +03:00
from libc.string cimport memset
2015-02-21 19:06:58 +03:00
def align(cand_words, gold_words):
cost, edit_path = _min_edit_path(cand_words, gold_words)
alignment = []
i_of_gold = 0
for move in edit_path:
if move == 'M':
alignment.append(i_of_gold)
i_of_gold += 1
elif move == 'S':
alignment.append(None)
i_of_gold += 1
elif move == 'D':
alignment.append(None)
elif move == 'I':
i_of_gold += 1
else:
raise Exception(move)
return alignment
punct_re = re.compile(r'\W')
def _min_edit_path(cand_words, gold_words):
cdef:
Pool mem
int i, j, n_cand, n_gold
int* curr_costs
int* prev_costs
# TODO: Fix this --- just do it properly, make the full edit matrix and
# then walk back over it...
mem = Pool()
# Preprocess inputs
cand_words = [punct_re.sub('', w) for w in cand_words]
gold_words = [punct_re.sub('', w) for w in gold_words]
n_cand = len(cand_words)
n_gold = len(gold_words)
# Levenshtein distance, except we need the history, and we may want different
# costs.
# Mark operations with a string, and score the history using _edit_cost.
previous_row = []
prev_costs = <int*>mem.alloc(n_gold + 1, sizeof(int))
curr_costs = <int*>mem.alloc(n_gold + 1, sizeof(int))
for i in range(n_gold + 1):
cell = ''
for j in range(i):
cell += 'I'
previous_row.append('I' * i)
prev_costs[i] = i
for i, cand in enumerate(cand_words):
current_row = ['D' * (i + 1)]
curr_costs[0] = i+1
for j, gold in enumerate(gold_words):
if gold.lower() == cand.lower():
s_cost = prev_costs[j]
i_cost = curr_costs[j] + 1
d_cost = prev_costs[j + 1] + 1
else:
s_cost = prev_costs[j] + 1
i_cost = curr_costs[j] + 1
d_cost = prev_costs[j + 1] + (1 if cand else 0)
if s_cost <= i_cost and s_cost <= d_cost:
best_cost = s_cost
best_hist = previous_row[j] + ('M' if gold == cand else 'S')
elif i_cost <= s_cost and i_cost <= d_cost:
best_cost = i_cost
best_hist = current_row[j] + 'I'
else:
best_cost = d_cost
best_hist = previous_row[j + 1] + 'D'
current_row.append(best_hist)
curr_costs[j+1] = best_cost
previous_row = current_row
for j in range(len(gold_words) + 1):
prev_costs[j] = curr_costs[j]
curr_costs[j] = 0
return prev_costs[n_gold], previous_row[-1]
2015-05-06 17:27:31 +03:00
def read_json_file(loc):
paragraphs = []
for doc in json.load(open(loc)):
for paragraph in doc['paragraphs']:
words = []
ids = []
tags = []
heads = []
labels = []
2015-05-27 04:17:50 +03:00
ner = []
2015-05-06 17:27:31 +03:00
for token in paragraph['tokens']:
words.append(token['orth'])
ids.append(token['id'])
2015-05-06 17:27:31 +03:00
tags.append(token['tag'])
heads.append(token['head'] if token['head'] >= 0 else token['id'])
2015-05-06 17:27:31 +03:00
labels.append(token['dep'])
2015-05-27 04:17:50 +03:00
ner.append(token.get('ner', '-'))
2015-05-06 17:27:31 +03:00
brackets = []
2015-05-27 04:17:50 +03:00
paragraphs.append((
paragraph['raw'],
(ids, words, tags, heads, labels, ner),
2015-05-11 17:12:03 +03:00
paragraph.get('brackets', [])))
2015-05-06 17:27:31 +03:00
return paragraphs
def _iob_to_biluo(tags):
out = []
curr_label = None
tags = list(tags)
while tags:
out.extend(_consume_os(tags))
out.extend(_consume_ent(tags))
return out
def _consume_os(tags):
while tags and tags[0] == 'O':
yield tags.pop(0)
def _consume_ent(tags):
if not tags:
return []
target = tags.pop(0).replace('B', 'I')
length = 1
while tags and tags[0] == target:
length += 1
tags.pop(0)
label = target[2:]
if length == 1:
return ['U-' + label]
else:
start = 'B-' + label
end = 'L-' + label
middle = ['I-%s' % label for _ in range(1, length - 1)]
return [start] + middle + [end]
2015-03-09 08:46:22 +03:00
cdef class GoldParse:
2015-05-12 23:33:47 +03:00
def __init__(self, tokens, annot_tuples, brackets=tuple()):
2015-03-09 08:46:22 +03:00
self.mem = Pool()
self.loss = 0
2015-02-23 22:04:53 +03:00
self.length = len(tokens)
2015-03-09 08:46:22 +03:00
# These are filled by the tagger/parser/entity recogniser
self.c_tags = <int*>self.mem.alloc(len(tokens), sizeof(int))
2015-03-09 08:46:22 +03:00
self.c_heads = <int*>self.mem.alloc(len(tokens), sizeof(int))
self.c_labels = <int*>self.mem.alloc(len(tokens), sizeof(int))
self.c_ner = <Transition*>self.mem.alloc(len(tokens), sizeof(Transition))
2015-05-11 17:12:03 +03:00
self.c_brackets = <int**>self.mem.alloc(len(tokens), sizeof(int*))
for i in range(len(tokens)):
self.c_brackets[i] = <int*>self.mem.alloc(len(tokens), sizeof(int))
2015-03-09 08:46:22 +03:00
self.tags = [None] * len(tokens)
self.heads = [None] * len(tokens)
self.labels = [''] * len(tokens)
self.ner = ['-'] * len(tokens)
self.cand_to_gold = align([t.orth_ for t in tokens], annot_tuples[1])
self.gold_to_cand = align(annot_tuples[1], [t.orth_ for t in tokens])
self.orig_annot = zip(*annot_tuples)
2015-03-09 08:46:22 +03:00
for i, gold_i in enumerate(self.cand_to_gold):
if gold_i is None:
# TODO: What do we do for missing values again?
2015-03-09 08:46:22 +03:00
pass
else:
self.tags[i] = annot_tuples[2][gold_i]
self.heads[i] = self.gold_to_cand[annot_tuples[3][gold_i]]
self.labels[i] = annot_tuples[4][gold_i]
2015-05-27 04:17:50 +03:00
self.ner[i] = annot_tuples[5][gold_i]
2015-05-11 17:12:03 +03:00
self.brackets = {}
for (gold_start, gold_end, label_str) in brackets:
start = self.gold_to_cand[gold_start]
end = self.gold_to_cand[gold_end]
if start is not None and end is not None:
self.brackets.setdefault(start, {}).setdefault(end, set())
2015-05-27 04:17:50 +03:00
self.brackets[end][start].add(label_str)
2015-05-11 17:12:03 +03:00
def __len__(self):
return self.length
2015-03-09 08:46:22 +03:00
2015-02-21 19:06:58 +03:00
def is_punct_label(label):
return label == 'P' or label.lower() == 'punct'