explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-11-04 09:57:26 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Oracle now consistent over gold standard derivation

Browse Source
This commit is contained in:
Matthew Honnibal 2015-01-29 03:41:58 +11:00
parent f590382134
commit 320b045daa
1 changed files with 8 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
spacy/syntax/arc_eager.pyx
View File

@ -58,9 +58,12 @@ cdef inline bint _can_break_shift(const State* s) nogil:
# P. 757 # P. 757
# In UPP, if Shift(F) or RightArc(F) fail to result in a single parsing # In UPP, if Shift(F) or RightArc(F) fail to result in a single parsing
# tree, they cannot be performed as well. # tree, they cannot be performed as well.
seen_headless = False
for i in range(s.stack_len): for i in range(s.stack_len):
if s.sent[s.stack[i]].head == 0: if seen_headless:
return False return False
else:
seen_headless = True
return True return True
@ -76,7 +79,7 @@ cdef inline bint _can_break_right(const State* s) nogil:
# tree, they cannot be performed as well. # tree, they cannot be performed as well.
seen_headless = False seen_headless = False
for i in range(s.stack_len): for i in range(s.stack_len):
if s.sent[s.stack[i]].head == 0: if s.sent[s.stack[-i]].head == 0:
if seen_headless: if seen_headless:
return False return False
else: else:
@ -123,6 +126,7 @@ cdef int _left_cost(const State* s, const int* gold) except -1:
cost += children_in_buffer(s, s.stack[0], gold) cost += children_in_buffer(s, s.stack[0], gold)
if NON_MONOTONIC and s.stack_len >= 2: if NON_MONOTONIC and s.stack_len >= 2:
cost += gold[s.stack[0]] == s.stack[-1] cost += gold[s.stack[0]] == s.stack[-1]
cost += gold[s.stack[0]] == s.stack[0]
return cost return cost
@ -140,18 +144,8 @@ cdef int _break_shift_cost(const State* s, const int* gold) except -1:
# #
# n0_cost: # n0_cost:
cdef int cost = 0 cdef int cost = 0
# number of head/child deps between n0 and N1...Nn # Number of deps between S0...Sn and N0...Nn
cost += children_in_buffer(s, s.i, gold) for i in range(s.i, s.sent_len):
cost += head_in_buffer(s, s.i, gold)
# Don't count self-deps
if gold[s.i] == s.i:
cost -= 2
# number of child deps from N0 into stack
cost += children_in_stack(s, s.i, gold)
# number of head deps to N0 from stack
cost += head_in_stack(s, s.i, gold)
# Number of deps between S0...Sn and N1...Nn
for i in range(s.i+1, s.sent_len):
cost += children_in_stack(s, i, gold) cost += children_in_stack(s, i, gold)
cost += head_in_stack(s, i, gold) cost += head_in_stack(s, i, gold)
return cost return cost
@ -255,8 +249,6 @@ cdef class TransitionSystem:
if not at_eol(s): if not at_eol(s):
push_stack(s) push_stack(s)
elif t.move == BREAK_SHIFT: elif t.move == BREAK_SHIFT:
push_stack(s)
get_s0(s).dep = 0
while s.stack_len != 0: while s.stack_len != 0:
s.stack -= 1 s.stack -= 1
s.stack_len -= 1 s.stack_len -= 1