explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2024-12-26 18:06:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Edits to arc eager oracle. Couldn't figure out how the non-monotonic lines made sense. They seem covered by children_in_stack

Browse Source
This commit is contained in:
Matthew Honnibal 2015-05-31 15:14:37 +02:00
parent 6c5632b71c
commit 5e99ff94c8
1 changed files with 5 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
spacy/syntax/arc_eager.pyx
View File

@ -238,8 +238,6 @@ cdef int _shift_cost(const Transition* self, const State* s, GoldParse gold) exc
cost = 0 cost = 0
cost += head_in_stack(s, s.i, gold.c_heads) cost += head_in_stack(s, s.i, gold.c_heads)
cost += children_in_stack(s, s.i, gold.c_heads) cost += children_in_stack(s, s.i, gold.c_heads)
if NON_MONOTONIC:
cost += gold.c_heads[s.stack[0]] == s.i
# If we can break, and there's no cost to doing so, we should # If we can break, and there's no cost to doing so, we should
if _can_break(s) and _break_cost(self, s, gold) == 0: if _can_break(s) and _break_cost(self, s, gold) == 0:
cost += 1 cost += 1
@ -258,8 +256,6 @@ cdef int _right_cost(const Transition* self, const State* s, GoldParse gold) exc
cost += head_in_buffer(s, s.i, gold.c_heads) cost += head_in_buffer(s, s.i, gold.c_heads)
cost += children_in_stack(s, s.i, gold.c_heads) cost += children_in_stack(s, s.i, gold.c_heads)
cost += head_in_stack(s, s.i, gold.c_heads) cost += head_in_stack(s, s.i, gold.c_heads)
if NON_MONOTONIC:
cost += gold.c_heads[s.stack[0]] == s.i
return cost return cost
@ -274,9 +270,11 @@ cdef int _left_cost(const Transition* self, const State* s, GoldParse gold) exce
elif at_eol(s): elif at_eol(s):
# Are we root? # Are we root?
if gold.c_labels[s.stack[0]] != -1: if gold.c_labels[s.stack[0]] != -1:
cost += gold.c_heads[s.stack[0]] != s.stack[0] # If we're at EOL, prefer to reduce or break over left-arc
# Are we labelling correctly? if _can_reduce(s) or _can_break(s):
cost += self.label != gold.c_labels[s.stack[0]] cost += gold.c_heads[s.stack[0]] != s.stack[0]
# Are we labelling correctly?
cost += self.label != gold.c_labels[s.stack[0]]
return cost return cost
cost += head_in_buffer(s, s.stack[0], gold.c_heads) cost += head_in_buffer(s, s.stack[0], gold.c_heads)