From 4662779251de3b692c20d4e10980a795f04e7520 Mon Sep 17 00:00:00 2001
From: Joe Zhao <ztuowen@gmail.com>
Date: Sun, 12 Apr 2015 10:59:08 +0800
Subject: validate, nDCG

---
 model/rankaccu.cpp | 69 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 model/rankaccu.h   | 13 ++++++++++
 2 files changed, 82 insertions(+)
 create mode 100644 model/rankaccu.cpp
 create mode 100644 model/rankaccu.h

(limited to 'model')

diff --git a/model/rankaccu.cpp b/model/rankaccu.cpp
new file mode 100644
index 0000000..2e77eb6
--- /dev/null
+++ b/model/rankaccu.cpp
@@ -0,0 +1,69 @@
+//
+// Created by joe on 4/12/15.
+//
+
+#include "rankaccu.h"
+#include "../tools/easylogging++.h"
+
+using namespace std;
+
+const double offset = 1;
+
+void ranksort(int l,int r,vector<int> &rank,const vector<double> &ref1,const vector<double> &ref2)
+{
+    int i=l,j=r,k;
+    double mid1=ref1[rank[(l+r)>>1]],mid2=ref2[rank[(l+r)>>1]];
+    while (i<=j)
+    {
+        while (ref1[rank[i]]>mid1 || (ref1[rank[i]]==mid1 && ref2[rank[i]]>mid2)) ++i;
+        while (ref1[rank[j]]<mid1 || (ref1[rank[j]]==mid1 && ref2[rank[j]]<mid2)) --j;
+        if (i<=j)
+        {
+            k=rank[i];
+            rank[i]=rank[j];
+            rank[j]=k;
+            ++i;
+            --j;
+        }
+    }
+    if (j>l)
+        ranksort(l,j,rank,ref1,ref2);
+    if (i<r)
+        ranksort(i,r,rank,ref1,ref2);
+}
+
+int rank_accu(DataList &D,const vector<double> pred)
+{
+    unsigned long n = D.getSize();
+    vector<int> orig_rank(n),pred_rank(n);
+    vector<double> orig(n);
+    int i,j;
+    for (i=0;i<D.getSize();++i)
+    {
+        orig_rank[i]=i;
+        pred_rank[i]=i;
+        orig[i]=D.getData()[i]->rank;
+    }
+    int cnt=0;
+    double accu_nDCG=0;
+    i=j=0;
+    while (i<D.getSize())
+    {
+        if ((i+1 == D.getSize())|| D.getData()[i]->qid!=D.getData()[i+1]->qid)
+        {
+            double Y=0,Z=0;
+            ranksort(j,i,orig_rank,orig,pred);
+            ranksort(j,i,pred_rank,pred,orig);
+            for (int k = j;k<=i;++k)
+            {
+                Z += (pow(2,offset+orig[orig_rank[k]]) - 1)/log2(2+k-j);
+                Y += (pow(2,offset+orig[pred_rank[k]]) - 1)/log2(2+k-j);
+            }
+            accu_nDCG+=Y/Z;
+            j = i+1;
+            ++cnt;
+        }
+        ++i;
+    }
+    LOG(INFO)<<"Average nDGC over "<< cnt<< " queries: "<< accu_nDCG/cnt;
+}
\ No newline at end of file
diff --git a/model/rankaccu.h b/model/rankaccu.h
new file mode 100644
index 0000000..3fe5379
--- /dev/null
+++ b/model/rankaccu.h
@@ -0,0 +1,13 @@
+//
+// Created by joe on 4/12/15.
+//
+
+#ifndef RANKSVM_RANKACCU_H
+#define RANKSVM_RANKACCU_H
+
+#include<vector>
+#include"../tools/dataProvider.h"
+
+int rank_accu(DataList &D,const std::vector<double> pred);
+
+#endif //RANKSVM_RANKACCU_H
-- 
cgit v1.2.3-70-g09d2