Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/moses-smt/mgiza.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
path: root/mgizapp/src/ATables.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'mgizapp/src/ATables.cpp')
-rw-r--r--mgizapp/src/ATables.cpp283
1 files changed, 144 insertions, 139 deletions
diff --git a/mgizapp/src/ATables.cpp b/mgizapp/src/ATables.cpp
index 8cc4ad2..7552e95 100644
--- a/mgizapp/src/ATables.cpp
+++ b/mgizapp/src/ATables.cpp
@@ -8,18 +8,18 @@ modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
-This program is distributed in the hope that it will be useful,
+This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
-Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
+Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
USA.
*/
-#include "ATables.h"
+#include "ATables.h"
#include "Globals.h"
#include "myassert.h"
#include "Parameter.h"
@@ -28,185 +28,190 @@ GLOBAL_PARAMETER(bool,CompactADTable,"compactadtable","1: only 3-dimensional ali
GLOBAL_PARAMETER(float,amodel_smooth_factor,"model23SmoothFactor","smoothing parameter for IBM-2/3 (interpolation with constant)",PARLEV_SMOOTH,0.0);
template <class VALTYPE>
-void amodel<VALTYPE>::printTable(const char *filename) const{
- // print amodel to file with the name filename (it'll be created or overwritten
- // format : for a table :
- // aj j l m val
- // where aj is source word pos, j target word pos, l source sentence length,
+void amodel<VALTYPE>::printTable(const char *filename) const
+{
+ // print amodel to file with the name filename (it'll be created or overwritten
+ // format : for a table :
+ // aj j l m val
+ // where aj is source word pos, j target word pos, l source sentence length,
// m is target sentence length.
- //
+ //
//return;
- if (is_distortion)
- cout << "Dumping pruned distortion table (d) to file:" << filename <<'\n';
- else
- cout << "Dumping pruned alignment table (a) to file:" << filename <<'\n';
-
- ofstream of(filename);
- double ssum=0.0;
- for(WordIndex l=0; l < MaxSentLength; l++){
- for(WordIndex m=0;m<MaxSentLength;m++){
- if( CompactADTable && l!=m )
- continue;
- unsigned int L=((CompactADTable&&is_distortion)?MaxSentLength:(l+1))-1;
- unsigned int M=((CompactADTable&&!is_distortion)?MaxSentLength:(m+1))-1;
- if( is_distortion==0 ){
- for(WordIndex j=1;j<=M; j++){
- double sum=0.0;
- for(WordIndex i=0;i<=L; i++){
- VALTYPE x=getValue(i, j, L, M);
- if( x>PROB_SMOOTH ){
- of << i << ' ' << j << ' ' << L << ' ' << M << ' ' << x << '\n';
- sum+=x;
- }
- }
- ssum+=sum;
- }
- }else{
- for(WordIndex i=0;i<=L;i++){
- double sum=0.0;
- for(WordIndex j=1;j<=M;j++){
- VALTYPE x=getValue(j, i, L, M);
- if( x>PROB_SMOOTH ){
- of << j << ' ' << i << ' ' << L << ' ' << M << ' ' << x << '\n';
- sum+=x;
- }
- }
- ssum+=sum;
- }
+ if (is_distortion)
+ cout << "Dumping pruned distortion table (d) to file:" << filename <<'\n';
+ else
+ cout << "Dumping pruned alignment table (a) to file:" << filename <<'\n';
+
+ ofstream of(filename);
+ double ssum=0.0;
+ for(WordIndex l=0; l < MaxSentLength; l++) {
+ for(WordIndex m=0; m<MaxSentLength; m++) {
+ if( CompactADTable && l!=m )
+ continue;
+ unsigned int L=((CompactADTable&&is_distortion)?MaxSentLength:(l+1))-1;
+ unsigned int M=((CompactADTable&&!is_distortion)?MaxSentLength:(m+1))-1;
+ if( is_distortion==0 ) {
+ for(WordIndex j=1; j<=M; j++) {
+ double sum=0.0;
+ for(WordIndex i=0; i<=L; i++) {
+ VALTYPE x=getValue(i, j, L, M);
+ if( x>PROB_SMOOTH ) {
+ of << i << ' ' << j << ' ' << L << ' ' << M << ' ' << x << '\n';
+ sum+=x;
+ }
+ }
+ ssum+=sum;
+ }
+ } else {
+ for(WordIndex i=0; i<=L; i++) {
+ double sum=0.0;
+ for(WordIndex j=1; j<=M; j++) {
+ VALTYPE x=getValue(j, i, L, M);
+ if( x>PROB_SMOOTH ) {
+ of << j << ' ' << i << ' ' << L << ' ' << M << ' ' << x << '\n';
+ sum+=x;
}
+ }
+ ssum+=sum;
}
+ }
}
+ }
}
template <class VALTYPE>
-void amodel<VALTYPE>::printRealTable(const char *filename) const{
- // print amodel to file with the name filename (it'll be created or overwritten
- // format : for a table :
- // aj j l m val
- // where aj is source word pos, j target word pos, l source sentence length,
+void amodel<VALTYPE>::printRealTable(const char *filename) const
+{
+ // print amodel to file with the name filename (it'll be created or overwritten
+ // format : for a table :
+ // aj j l m val
+ // where aj is source word pos, j target word pos, l source sentence length,
// m is target sentence length.
- //
+ //
//return;
- if (is_distortion)
- cout << "Dumping not pruned distortion table (d) to file:" << filename <<'\n';
- else
- cout << "Dumping not pruned alignment table (a) to file:" << filename <<'\n';
-
- ofstream of(filename);
- for(WordIndex l=0; l < MaxSentLength; l++){
- for(WordIndex m=0;m<MaxSentLength;m++){
- if( CompactADTable && l!=m )
- continue;
- unsigned int L=((CompactADTable&&is_distortion)?MaxSentLength:(l+1))-1;
- unsigned int M=((CompactADTable&&!is_distortion)?MaxSentLength:(m+1))-1;
- if( is_distortion==0 ){
- for(WordIndex j=1;j<=M; j++){
- for(WordIndex i=0;i<=L; i++){
- VALTYPE x=getValue(i, j, L, M);
- if( x>MINCOUNTINCREASE )
- of << i << ' ' << j << ' ' << L << ' ' << M << ' ' << x << '\n';
- }
- }
- }else{
- for(WordIndex i=0;i<=L;i++){
- for(WordIndex j=1;j<=M;j++){
- VALTYPE x=getValue(j, i, L, M);
- if( x>MINCOUNTINCREASE )
- of << j << ' ' << i << ' ' << L << ' ' << M << ' ' << x << '\n';
- }
- }
- }
+ if (is_distortion)
+ cout << "Dumping not pruned distortion table (d) to file:" << filename <<'\n';
+ else
+ cout << "Dumping not pruned alignment table (a) to file:" << filename <<'\n';
+
+ ofstream of(filename);
+ for(WordIndex l=0; l < MaxSentLength; l++) {
+ for(WordIndex m=0; m<MaxSentLength; m++) {
+ if( CompactADTable && l!=m )
+ continue;
+ unsigned int L=((CompactADTable&&is_distortion)?MaxSentLength:(l+1))-1;
+ unsigned int M=((CompactADTable&&!is_distortion)?MaxSentLength:(m+1))-1;
+ if( is_distortion==0 ) {
+ for(WordIndex j=1; j<=M; j++) {
+ for(WordIndex i=0; i<=L; i++) {
+ VALTYPE x=getValue(i, j, L, M);
+ if( x>MINCOUNTINCREASE )
+ of << i << ' ' << j << ' ' << L << ' ' << M << ' ' << x << '\n';
+ }
+ }
+ } else {
+ for(WordIndex i=0; i<=L; i++) {
+ for(WordIndex j=1; j<=M; j++) {
+ VALTYPE x=getValue(j, i, L, M);
+ if( x>MINCOUNTINCREASE )
+ of << j << ' ' << i << ' ' << L << ' ' << M << ' ' << x << '\n';
+ }
}
+ }
}
+ }
}
extern short NoEmptyWord;
template <class VALTYPE>
-bool amodel<VALTYPE>::readTable(const char *filename){
+bool amodel<VALTYPE>::readTable(const char *filename)
+{
/* This function reads the a table from a file.
Each line is of the format: aj j l m val
- where aj is the source word position, j the target word position,
+ where aj is the source word position, j the target word position,
l the source sentence length, and m the target sentence length
-
+
This function also works for a d table, where the positions
of aj and i are swapped. Both the a and d tables are 4 dimensional
hashes; this function will simply read in the four values and keep
them in that order when hashing the fifth value.
NAS, 7/11/99
*/
- ifstream inf(filename);
- cout << "Reading a/d table from " << filename << "\n";
- if(!inf){
- cerr << "\nERROR: Cannot open " << filename<<"\n";
- return false;
- }
- WordIndex w, x, l, m;
- VALTYPE prob;
- while(inf >> w >> x >> l >> m >> prob )
- // the NULL word is added to the length
- // of the sentence in the tables, but discount it when you write the tables.
- setValue(w, x, l, m, prob);
- return true;
+ ifstream inf(filename);
+ cout << "Reading a/d table from " << filename << "\n";
+ if(!inf) {
+ cerr << "\nERROR: Cannot open " << filename<<"\n";
+ return false;
+ }
+ WordIndex w, x, l, m;
+ VALTYPE prob;
+ while(inf >> w >> x >> l >> m >> prob )
+ // the NULL word is added to the length
+ // of the sentence in the tables, but discount it when you write the tables.
+ setValue(w, x, l, m, prob);
+ return true;
}
template <class VALTYPE>
-bool amodel<VALTYPE>::readAugTable(const char *filename){
+bool amodel<VALTYPE>::readAugTable(const char *filename)
+{
/* This function reads the a table from a file.
Each line is of the format: aj j l m val
- where aj is the source word position, j the target word position,
+ where aj is the source word position, j the target word position,
l the source sentence length, and m the target sentence length
-
+
This function also works for a d table, where the positions
of aj and i are swapped. Both the a and d tables are 4 dimensional
hashes; this function will simply read in the four values and keep
them in that order when hashing the fifth value.
NAS, 7/11/99
*/
- ifstream inf(filename);
- cout << "Reading a/d table from " << filename << "\n";
- if(!inf){
- cerr << "\nERROR: Cannot open " << filename<<"\n";
- return false;
- }
- WordIndex w, x, l, m;
- VALTYPE prob;
- while(inf >> w >> x >> l >> m >> prob )
- // the NULL word is added to the length
- // of the sentence in the tables, but discount it when you write the tables.
- addValue(w, x, l, m, prob);
- return true;
+ ifstream inf(filename);
+ cout << "Reading a/d table from " << filename << "\n";
+ if(!inf) {
+ cerr << "\nERROR: Cannot open " << filename<<"\n";
+ return false;
+ }
+ WordIndex w, x, l, m;
+ VALTYPE prob;
+ while(inf >> w >> x >> l >> m >> prob )
+ // the NULL word is added to the length
+ // of the sentence in the tables, but discount it when you write the tables.
+ addValue(w, x, l, m, prob);
+ return true;
}
template <class VALTYPE>
-bool amodel<VALTYPE>::merge(amodel<VALTYPE>& am){
- cout << "start merging " <<"\n";
- for(WordIndex l=0; l < MaxSentLength; l++){
- for(WordIndex m=0;m<MaxSentLength;m++){
- if( CompactADTable && l!=m )
- continue;
- unsigned int L=((CompactADTable&&is_distortion)?MaxSentLength:(l+1))-1;
- unsigned int M=((CompactADTable&&!is_distortion)?MaxSentLength:(m+1))-1;
- if( is_distortion==0 ){
- for(WordIndex j=1;j<=M; j++){
- for(WordIndex i=0;i<=L; i++){
- VALTYPE x=am.getValue(i, j, L, M);
- addValue(i,j,L,M,x);
- }
- }
- }else{
- for(WordIndex i=0;i<=L;i++){
- for(WordIndex j=1;j<=M;j++){
- VALTYPE x=am.getValue(j, i, L, M);
- addValue(j,i,L,M,x);
- }
- }
- }
+bool amodel<VALTYPE>::merge(amodel<VALTYPE>& am)
+{
+ cout << "start merging " <<"\n";
+ for(WordIndex l=0; l < MaxSentLength; l++) {
+ for(WordIndex m=0; m<MaxSentLength; m++) {
+ if( CompactADTable && l!=m )
+ continue;
+ unsigned int L=((CompactADTable&&is_distortion)?MaxSentLength:(l+1))-1;
+ unsigned int M=((CompactADTable&&!is_distortion)?MaxSentLength:(m+1))-1;
+ if( is_distortion==0 ) {
+ for(WordIndex j=1; j<=M; j++) {
+ for(WordIndex i=0; i<=L; i++) {
+ VALTYPE x=am.getValue(i, j, L, M);
+ addValue(i,j,L,M,x);
+ }
+ }
+ } else {
+ for(WordIndex i=0; i<=L; i++) {
+ for(WordIndex j=1; j<=M; j++) {
+ VALTYPE x=am.getValue(j, i, L, M);
+ addValue(j,i,L,M,x);
+ }
}
+ }
}
- return true;
+ }
+ return true;
}
-template class amodel<COUNT> ;
-//template class amodel<PROB> ;
+template class amodel<COUNT> ;
+//template class amodel<PROB> ;
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-18 05:01:14 +0300