bug

sdk/Evaluation.cpp

@@ -5,14 +5,13 @@
 
 struct minEval{
         double value;
-        double values;
+        int* values;
 
 	double eval;
         long left; // how many on its left
         double* record;
         long max;
         long** count;
-        long* sorted; // sorted d
 };
 
 minEval giniSparse(double** data, long* result, long* d, long size, long col, long classes, long* totalT){

sdk/irf.cpp

@@ -4,9 +4,6 @@
 #include "../server/table.h"
 
 DecisionTree* dt = nullptr;
-long pt = 0;
-double** data = nullptr;
-long* result = nullptr;
 
 __AQEXPORT__(bool) newtree(int height, long f, ColRef<int> sparse, double forget, long maxf, long noclasses, Evaluation e, long r, long rb){
 	if(sparse.size!=f)return 0;
@@ -19,36 +16,29 @@ __AQEXPORT__(bool) newtree(int height, long f, ColRef<int> sparse, double forget
 	return 1;
 }
 
-__AQEXPORT__(bool) additem(ColRef<double>X, long y, long size){
+__AQEXPORT__(bool) fit(ColRef<ColRef<double>> X, ColRef<int> y){
-	long j = 0;
+	if(X.size != y.size)return 0;
-	if(size>0){
+	double** data = (double**)malloc(X.size*sizeof(double*));
-		free(data);
+	long* result = (long*)malloc(y.size*sizeof(long));
-		free(result);
+	for(long i=0; i<X.size; i++){
-		pt = 0;
+		data[i] = X.container[i].container;
-		data=(double**)malloc(size*sizeof(double*));
+		result[i] = y.container[i];
-		result=(long*)malloc(size*sizeof(long));
+	}
-	}
+	dt->fit(data, result, X.size);
-	data[pt] = (double*)malloc(X.size*sizeof(double));
-	for(j=0; j<X.size; j++){
-		data[pt][j]=X.container[j];
-	}
-	result[pt] = y;
-	pt ++;
-	return 1;
-}
-__AQEXPORT__(bool) fit(){
-	if(pt<=0)return 0;
-	dt->fit(data, result, pt);
 	return 1;
 }
 
-__AQEXPORT__(ColRef_storage) predict(){
+__AQEXPORT__(ColRef_storage) predict(ColRef<ColRef<double>> X){
-	int* result = (int*)malloc(pt*sizeof(int));
+        double** data = (double**)malloc(X.size*sizeof(double*));
-	for(long i=0; i<pt; i++){
+        int* result = (int*)malloc(X.size*sizeof(int));
+        for(long i=0; i<X.size; i++){
+                data[i] = X.container[i].container;
+        }
+	for(long i=0; i<X.size; i++){
 		result[i]=dt->Test(data[i], dt->DTree);
 	}
 	
-	return ColRef_storage(new ColRef_storage(result, pt, 0, "prediction", 0), 1, 0, "prediction", 0);
+	return ColRef_storage(new ColRef_storage(result, X.size, 0, "prediction", 0), 1, 0, "prediction", 0);
 }
 
 

tests/dt.a

@@ -1,21 +1,21 @@
 LOAD MODULE FROM "./libirf.so"
 FUNCTIONS (
     newtree(height:int, f:int64, sparse:vecint, forget:double, maxf:int64, noclasses:int64, e:int, r:int64, rb:int64) -> bool,
-    additem(X:vecdouble, y:int64, size:int64) -> bool,
+    fit(X:vecvecdouble, y:vecint) -> bool,
-    fit() -> bool,
+    predict(X:vecvecdouble) -> vecint
-    predict() -> vecint
 );
-create table tb(x int);
+
-create table tb2(x double, y double, z double);
+create table source(x1 double, x2 double, x3 double, x4 double, x5 int);
-insert into tb values (0);
+load data infile "data/benchmark" into table source fields terminated by ",";
-insert into tb values (0);
+
-insert into tb values (0);
+create table sparse(x int);
-select newtree(5, 3, tb.x, 0, 3, 2, 0, 100, 1) from tb;
+insert into sparse values (1);
-insert into tb2 values (1, 0, 1);
+insert into sparse values (1);
-insert into tb2 values (0, 1, 1);
+insert into sparse values (1);
-insert into tb2 values (1, 1, 1);
+insert into sparse values (1);
-select additem(tb2.x, 1, 3) from tb2;
+
-select additem(tb2.y, 0, -1) from tb2;
+select newtree(6, 4, sparse.x, 0, 4, 2, 0, 400, 2147483647) from sparse;
-select additem(tb2.z, 1, -1) from tb2;
+
-select fit();
+select fit(pack(x1, x2, x3, x4), x5) from source;
-select predict();
+
+select predict(pack(x1, x2, x3, x4)) from source;