From 3c1c98b5aec7aec34432ddc37385df06d301bdd5 Mon Sep 17 00:00:00 2001
From: Chris Dyer <cdyer@cs.cmu.edu>
Date: Fri, 27 Jan 2012 02:31:00 -0500
Subject: migrate mert to the new scorer interface

---
 vest/mr_vest_reduce.cc | 34 +++++++++++++++-------------------
 1 file changed, 15 insertions(+), 19 deletions(-)

(limited to 'vest/mr_vest_reduce.cc')

diff --git a/vest/mr_vest_reduce.cc b/vest/mr_vest_reduce.cc
index 3df52020..dda61f88 100644
--- a/vest/mr_vest_reduce.cc
+++ b/vest/mr_vest_reduce.cc
@@ -10,6 +10,7 @@
 #include "error_surface.h"
 #include "line_optimizer.h"
 #include "b64tools.h"
+#include "stringlib.h"
 
 using namespace std;
 namespace po = boost::program_options;
@@ -17,12 +18,12 @@ namespace po = boost::program_options;
 void InitCommandLine(int argc, char** argv, po::variables_map* conf) {
   po::options_description opts("Configuration options");
   opts.add_options()
-        ("loss_function,l",po::value<string>(), "Loss function being optimized")
+        ("evaluation_metric,m",po::value<string>(), "Evaluation metric (IBM_BLEU, etc.)")
         ("help,h", "Help");
   po::options_description dcmdline_options;
   dcmdline_options.add(opts);
   po::store(parse_command_line(argc, argv, dcmdline_options), *conf);
-  bool flag = conf->count("loss_function") == 0;
+  bool flag = conf->count("evaluation_metric") == 0;
   if (flag || conf->count("help")) {
     cerr << dcmdline_options << endl;
     exit(1);
@@ -32,30 +33,27 @@ void InitCommandLine(int argc, char** argv, po::variables_map* conf) {
 int main(int argc, char** argv) {
   po::variables_map conf;
   InitCommandLine(argc, argv, &conf);
-  const string loss_function = conf["loss_function"].as<string>();
-  ScoreType type = ScoreTypeFromString(loss_function);
+  const string evaluation_metric = conf["evaluation_metric"].as<string>();
   LineOptimizer::ScoreType opt_type = LineOptimizer::MAXIMIZE_SCORE;
-  if (type == TER || type == AER) {
+  if (UppercaseString(evaluation_metric) == "TER")
     opt_type = LineOptimizer::MINIMIZE_SCORE;
-  }
-  string last_key;
+  EvaluationMetric* metric = EvaluationMetric::Instance(evaluation_metric);
+
   vector<ErrorSurface> esv;
-  while(cin) {
-    string line;
-    getline(cin, line);
-    if (line.empty()) continue;
+  string last_key, line, key, val;
+  while(getline(cin, line)) {
     size_t ks = line.find("\t");
     assert(string::npos != ks);
     assert(ks > 2);
-    string key = line.substr(2, ks - 2);
-    string val = line.substr(ks + 1);
+    key = line.substr(2, ks - 2);
+    val = line.substr(ks + 1);
     if (key != last_key) {
       if (!last_key.empty()) {
 	float score;
-        double x = LineOptimizer::LineOptimize(esv, opt_type, &score);
+        double x = LineOptimizer::LineOptimize(metric, esv, opt_type, &score);
 	cout << last_key << "|" << x << "|" << score << endl;
       }
-      last_key = key;
+      last_key.swap(key);
       esv.clear();
     }
     if (val.size() % 4 != 0) {
@@ -68,13 +66,11 @@ int main(int argc, char** argv) {
       continue;
     }
     esv.push_back(ErrorSurface());
-    esv.back().Deserialize(type, encoded);
+    esv.back().Deserialize(encoded);
   }
   if (!esv.empty()) {
-    // cerr << "ESV=" << esv.size() << endl;
-    // for (int i = 0; i < esv.size(); ++i) { cerr << esv[i].size() << endl; }
     float score;
-    double x = LineOptimizer::LineOptimize(esv, opt_type, &score);
+    double x = LineOptimizer::LineOptimize(metric, esv, opt_type, &score);
     cout << last_key << "|" << x << "|" << score << endl;
   }
   return 0;
-- 
cgit v1.2.3