Test for Dense Logistic

ML10/ML10_Logistic.ClassifyD_UsageExample.ecl

@@ -1,4 +1,5 @@
 //ML10_Logistic.ClassifyD_UsageExample.ecl
+//Uses sparse implementation of Logistic
 IMPORT ML;
 value_record := RECORD
 UNSIGNED rid;
@@ -16,17 +17,37 @@ d := DATASET([{1,35,149,0},{2,11,138,0},{3,12,148,1},{4,16,156,0},
 {29,20,162,0},{30,25,180,1},{31,22,173,1},{32,25,171,1}]
 ,value_record);
 OUTPUT(d,NAMED('d'));
+
 ML.ToField(d,flds0);
 OUTPUT(flds0,NAMED('flds0'));
+
 flds := ML.Discretize.ByRounding(flds0);
 OUTPUT(SORT(flds,id,number),NAMED('flds'));
-MyLogisticRegression:=ML.Classify.Logistic();
-Model3 := MyLogisticRegression.LearnD(flds(Number<=2),flds(Number=3));
-OUTPUT(Model3,NAMED('Model3'));
+
+MyLogisticRegression:=ML.Classify.Logistic_sparse();
+
+Model3 := MyLogisticRegression.LearnD(flds(Number<=2),flds(Number>=3));
+OUTPUT(SORT(Model3,id,number),NAMED('Model3'));
+
 predict:=MyLogisticRegression.ClassifyD(flds(Number<=2),Model3);
-OUTPUT(predict,NAMED('predict'));
+OUTPUT(SORT(predict,id,number),NAMED('predict'));
+
+predicted_v_actual :=
+  JOIN(predict,flds(Number>=3)
+       ,(LEFT.id=RIGHT.id) and (LEFT.number=RIGHT.number)
+       ,TRANSFORM({UNSIGNED id,UNSIGNED4 number,REAL8 pvalue,REAL8 avalue}
+                  ,SELF.pvalue:=LEFT.value
+                  ,SELF.avalue:=RIGHT.value
+                  ,SELF := LEFT
+        )
+  );
+OUTPUT(SORT(predicted_v_actual,id,number),NAMED('predicted_v_actual'));
+
+Accuracy_proportion := COUNT(predicted_v_actual(pvalue=avalue))/COUNT(predicted_v_actual);
+Output(Accuracy_proportion, NAMED('Accuracy'));
+
 TestModule := MyLogisticRegression.TestD(flds(Number<=2),flds(Number=3));
 TestModule.Raw;
 TestModule.CrossAssignments;
 TestModule.PrecisionByClass;
-TestModule.Headline;
+TestModule.Headline;

ML10/ML10_LogisticDense.ClassifyD_UsageExample.ecl

@@ -0,0 +1,64 @@
+//ML10_LogisticDense.ClassifyD_UsageExample2.ecl
+IMPORT ML;
+value_record := RECORD
+UNSIGNED rid;
+REAL age;
+REAL height;
+integer1 sex; // 0 = female, 1 = male
+END;
+d := DATASET([{1,35,149,0},{2,11,138,0},{3,12,148,1},{4,16,156,0},
+{5,32,152,0},{6,16,157,0},{7,14,165,0},{8,8,152,1},
+{9,35,177,0},{10,33,158,1},{11,40,166,0},{12,28,165,0},
+{13,23,160,0},{14,52,178,1},{15,46,169,0},{16,29,173,1},
+{17,30,172,0},{18,21,163,0},{19,21,164,0},{20,20,189,1},
+{21,34,182,1},{22,43,184,1},{23,35,174,1},{24,39,177,1},
+{25,43,183,1},{26,37,175,1},{27,32,173,1},{28,24,173,1},
+{29,20,162,0},{30,25,180,1},{31,22,173,1},{32,25,171,1}]
+,value_record);
+OUTPUT(d,NAMED('d'));
+
+ML.ToField(d,flds0);
+OUTPUT(flds0,NAMED('flds0'));
+
+flds := ML.Discretize.ByRounding(flds0);
+OUTPUT(SORT(flds,id,number),NAMED('flds'));
+
+//Uses PBblas/AutoMap.ecl
+MyLogisticRegression:=ML.Classify.Logistic();
+
+
+//Manually sets partition block sizes
+
+/*Note that though there are to independent
+  variables, the partition size for the columns
+  is 3. This is because a column of 1's is 
+  inserted in the Logistic module.
+*/
+
+//MyLogisticRegression:=ML.Classify.Logistic(,,,32,3);
+
+Model3 := MyLogisticRegression.LearnD(flds(Number<=2),flds(Number>=3));
+OUTPUT(SORT(Model3,id,number),NAMED('Model3'));
+
+predict:=MyLogisticRegression.ClassifyD(flds(Number<=2),Model3);
+OUTPUT(SORT(predict,id,number),NAMED('predict'));
+
+predicted_v_actual :=
+  JOIN(predict,flds(Number>=3)
+       ,(LEFT.id=RIGHT.id) and (LEFT.number=RIGHT.number)
+       ,TRANSFORM({UNSIGNED id,UNSIGNED4 number,REAL8 pvalue,REAL8 avalue}
+                  ,SELF.pvalue:=LEFT.value
+                  ,SELF.avalue:=RIGHT.value
+                  ,SELF := LEFT
+        )
+  );
+OUTPUT(SORT(predicted_v_actual,id,number),NAMED('predicted_v_actual'));
+
+Accuracy_proportion := COUNT(predicted_v_actual(pvalue=avalue))/COUNT(predicted_v_actual);
+Output(Accuracy_proportion, NAMED('Accuracy'));
+
+TestModule := MyLogisticRegression.TestD(flds(Number<=2),flds(Number=3));
+TestModule.Raw;
+TestModule.CrossAssignments;
+TestModule.PrecisionByClass;
+TestModule.Headline;