summaryrefslogtreecommitdiff
path: root/word-aligner/makefiles
diff options
context:
space:
mode:
authorredpony <redpony@ec762483-ff6d-05da-a07a-a48fb63a330f>2010-12-01 05:27:13 +0000
committerredpony <redpony@ec762483-ff6d-05da-a07a-a48fb63a330f>2010-12-01 05:27:13 +0000
commitd52db01a2e224869c6ea72a4a234e888c6fd756c (patch)
treec5aff0967b4fcca2ac879aecb4ac68317d3582aa /word-aligner/makefiles
parentca83fbe4c043b2e4e18a21f91e74dfa922eda44e (diff)
alternative def of neighborhoods
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@739 ec762483-ff6d-05da-a07a-a48fb63a330f
Diffstat (limited to 'word-aligner/makefiles')
-rw-r--r--word-aligner/makefiles/makefile.grammars22
1 files changed, 16 insertions, 6 deletions
diff --git a/word-aligner/makefiles/makefile.grammars b/word-aligner/makefiles/makefile.grammars
index 21f39ac1..60417ec5 100644
--- a/word-aligner/makefiles/makefile.grammars
+++ b/word-aligner/makefiles/makefile.grammars
@@ -1,7 +1,7 @@
-all: corpus.f-e.lex-grammar.gz corpus.class.e corpus.class.f corpus.stemmed.f fstem.map corpus.stemmed.e estem.map corpus.f-e.sgml
+all: corpus.f-e.lex-grammar.gz wordpairs.f-e.features.gz corpus.class.e corpus.class.f corpus.stemmed.f fstem.map corpus.stemmed.e estem.map corpus.f-e.sgml
clean:
- $(RM) orthonorm-dict.* voc2class* corpus.class.* corpus.e-f corpus.f-e corpus.f-e.lex-grammar* *.model1 *voc corpus.e-f.lex-grammar* *stem* corpus.f-e.sgml freq* psg*
+ $(RM) orthonorm-dict.* voc2class* corpus.class.* corpus.e-f corpus.f-e corpus.f-e.lex-grammar* *.model1 *voc corpus.e-f.lex-grammar* *stem* corpus.f-e.sgml freq* psg* wordpairs*
SUPPORT_DIR = $(SCRIPT_DIR)/support
GZIP = /usr/bin/gzip
@@ -9,6 +9,7 @@ ZCAT = zcat
EXTRACT_GRAMMAR = $(SUPPORT_DIR)/extract_grammar.pl
EXTRACT_VOCAB = $(SUPPORT_DIR)/extract_vocab.pl
GENERATE_PSG = $(SUPPORT_DIR)/generate_per_sentence_grammars.pl
+GENERATE_WORDPAIR_FEATURES = $(SUPPORT_DIR)/generate_word_pair_features.pl
ORTHONORM_E = $(SCRIPT_DIR)/ortho-norm/$(E_LANG).pl
ORTHONORM_F = $(SCRIPT_DIR)/ortho-norm/$(F_LANG).pl
STEM_F = $(SCRIPT_DIR)/stemmers/$(F_LANG).pl
@@ -66,13 +67,22 @@ corpus.e-f: corpus.f corpus.e $(MERGE_CORPUS)
$(MERGE_CORPUS) corpus.e corpus.f > $@
corpus.f-e.model1: corpus.f-e $(MODEL1)
- $(MODEL1) corpus.f-e > $@
+ $(MODEL1) -v -V corpus.f-e > $@
corpus.e-f.model1: corpus.e-f $(MODEL1)
- $(MODEL1) corpus.e-f > $@
+ $(MODEL1) -v -V corpus.e-f > $@
-corpus.f-e.lex-grammar.gz: corpus.f-e corpus.f-e.model1 corpus.e-f.model1 orthonorm-dict.f orthonorm-dict.e voc2class.e voc2class.f $(GIZAALIGN) $(INVGIZAALIGN)
- $(MAKE_LEX_GRAMMAR) corpus.f-e corpus.f-e.model1 corpus.e-f.model1 orthonorm-dict.f orthonorm-dict.e voc2class.e voc2class.f $(GIZAALIGN) $(INVGIZAALIGN) | $(GZIP) -9 > $@
+corpus.f-e.full-model1: corpus.f-e $(MODEL1)
+ $(MODEL1) -t -999999 -v -V corpus.f-e > $@
+
+corpus.e-f.full-model1: corpus.e-f $(MODEL1)
+ $(MODEL1) -t -999999 -v -V corpus.e-f > $@
+
+corpus.f-e.lex-grammar.gz: corpus.f-e corpus.f-e.model1 corpus.e-f.model1
+ $(MAKE_LEX_GRAMMAR) corpus.f-e corpus.f-e.model1 corpus.e-f.model1 | $(GZIP) -9 > $@
+
+wordpairs.f-e.features.gz: corpus.f-e corpus.f-e.full-model1 corpus.e-f.full-model1 orthonorm-dict.f orthonorm-dict.e voc2class.e voc2class.f corpus.f-e.model1
+ $(GENERATE_WORDPAIR_FEATURES) corpus.f-e corpus.f-e.full-model1 corpus.e-f.full-model1 orthonorm-dict.f orthonorm-dict.e voc2class.e voc2class.f corpus.f-e.model1 | $(GZIP) -9 > $@
corpus.f-e.sgml: f.voc corpus.f-e.lex-grammar.gz corpus.f-e
$(GENERATE_PSG) f.voc corpus.f-e corpus.f-e.lex-grammar.gz freq_grammar.f-e.gz psg.f-e $@