From e65f1c2448b0970d9ef296d682a7f862c7663ffb Mon Sep 17 00:00:00 2001
From: rylon <rylon@web.de>
Date: Sat, 15 Jun 2013 23:00:45 +0200
Subject: [PATCH] webmining u4 update

---
 .../4_Uebung/code/confusion_matrix.py         | 39 ++++++++++++-------
 1 file changed, 24 insertions(+), 15 deletions(-)

diff --git a/ss2013/1_Web Mining/Uebungen/4_Uebung/code/confusion_matrix.py b/ss2013/1_Web Mining/Uebungen/4_Uebung/code/confusion_matrix.py
index 55537129..5580c2f5 100644
--- a/ss2013/1_Web Mining/Uebungen/4_Uebung/code/confusion_matrix.py	
+++ b/ss2013/1_Web Mining/Uebungen/4_Uebung/code/confusion_matrix.py	
@@ -54,6 +54,7 @@ def accuracy():
     accuracy = float(ok_recognized) / float(document_count)
     
 def prec():
+    #per class -> positive matches / alle matches auf class
     i = 0
     global precision
     for conf in conf_matr:
@@ -69,7 +70,7 @@ def prec():
         if not_ok_values +ok_values > 0:
             precision[i] = float(ok_values) / float(ok_values + not_ok_values)
         else:
-            precision[i] = 0
+            precision[i] = 0 #division by zero
         i += 1
 
 def avg_prec():
@@ -81,26 +82,34 @@ def avg_prec():
 
        
 def pp_confusionmatrix():
-    print "\tc1\tc2\tc3\tc4\tc5\tc6\tc7\tc8\tc9\tc10\tc11"
-    lines = ["c1\t","c2\t","c3\t","c4\t","c5\t","c6\t","c7\t","c8\t","c9\t","c10\t","c11\t"]
+    print "\t\t"+classes[0][0]+"\t"+classes[1][0]+"\t"+classes[2][0]+"\t"+classes[3][0]+"\t"+classes[4][0]+"\t"+classes[5][0]+"\t"+classes[6][0]+"\t"+classes[7][0]+"\t"+classes[8][0]+"\t"+classes[9][0]+"\t"+classes[10][0]
+    lines = [""+classes[0][0]+"\t",classes[1][0]+"\t",""+classes[2][0]+"\t",classes[3][0]+"\t\t",classes[4][0]+"\t",classes[5][0]+"\t\t",classes[6][0]+"\t\t",classes[7][0]+"\t\t",classes[8][0]+"\t\t",classes[9][0]+"\t\t",classes[10][0]+"\t\t"]
     for column in conf_matr:
-        lines[0] += str(column[0]) + "\t"
-        lines[1] += str(column[1]) + "\t"
-        lines[2] += str(column[2]) + "\t"
-        lines[3] += str(column[3]) + "\t"
-        lines[4] += str(column[4]) + "\t"
-        lines[5] += str(column[5]) + "\t"
-        lines[6] += str(column[6]) + "\t"
-        lines[7] += str(column[7]) + "\t"
-        lines[8] += str(column[8]) + "\t"
-        lines[9] += str(column[9]) + "\t"
-        lines[10] += str(column[10]) + "\t"
+        lines[0] += str(column[0]) + "\t\t"
+        lines[1] += str(column[1]) + "\t\t"
+        lines[2] += str(column[2]) + "\t\t"
+        lines[3] += str(column[3]) + "\t\t"
+        lines[4] += str(column[4]) + "\t\t"
+        lines[5] += str(column[5]) + "\t\t"
+        lines[6] += str(column[6]) + "\t\t"
+        lines[7] += str(column[7]) + "\t\t"
+        lines[8] += str(column[8]) + "\t\t"
+        lines[9] += str(column[9]) + "\t\t"
+        lines[10] += str(column[10]) + "\t\t"
 
     for l in lines:
         print l
 
 def pp_accuracy():
     print "Accuracy: "+str(round(accuracy*100,4))+"%"
+    
+def pp_prec():
+    line = "Precision per class: "
+    i = 0
+    for p in precision:
+        line += classes[i][0]+":"+str(round(p*100,4))+" "
+        i += 1
+    print line
 
 if __name__ == '__main__':
     load_predictionfile(predfile)
@@ -111,7 +120,7 @@ if __name__ == '__main__':
     accuracy()
     pp_accuracy()
     prec()
-    print precision
+    pp_prec()
     avg_prec()
     print precision_avg
     
\ No newline at end of file