complete svm-train & svm-predict

Xtra-Computing · Nov 2, 2017 · a7fb32c · a7fb32c
1 parent 05b6cea
commit a7fb32c
Show file tree

Hide file tree

Showing 21 changed files with 117 additions and 195 deletions.
diff --git a/include/thundersvm/dataset.h b/include/thundersvm/dataset.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_DATASET_H
 #define THUNDERSVM_DATASET_H
 
-#include "thundersvm-train.h"
+#include "thundersvm.h"
 #include "syncdata.h"
 class DataSet {
 public:

diff --git a/include/thundersvm/kernel/kernelmatrix_kernel.h b/include/thundersvm/kernel/kernelmatrix_kernel.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_KERNELMATRIX_KERNEL_H
 #define THUNDERSVM_KERNELMATRIX_KERNEL_H
 
-#include "thundersvm/thundersvm-train.h"
+#include "thundersvm/thundersvm.h"
 #include "thundersvm/clion_cuda.h"
 
 __global__ void

diff --git a/include/thundersvm/kernel/smo_kernel.h b/include/thundersvm/kernel/smo_kernel.h
@@ -6,7 +6,7 @@
 #define THUNDERSVM_SMO_KERNEL_H
 
 #include <thundersvm/clion_cuda.h>
-#include <thundersvm/thundersvm-train.h>
+#include <thundersvm/thundersvm.h>
 
 __host__ __device__
 

diff --git a/include/thundersvm/kernelmatrix.h b/include/thundersvm/kernelmatrix.h
@@ -6,7 +6,7 @@
 #define THUNDERSVM_KERNELMATRIX_H
 
 #include <cusparse.h>
-#include "thundersvm-train.h"
+#include "thundersvm.h"
 #include "syncdata.h"
 #include "dataset.h"
 #include "svmparam.h"

diff --git a/include/thundersvm/model/oneclass_svc.h b/include/thundersvm/model/oneclass_svc.h
@@ -11,8 +11,6 @@ class OneClassSVC : public SvmModel {
 public:
     void train(DataSet dataset, SvmParam param) override;
 
-//    void load_from_file(string path) override;
-
     vector<real> predict(const DataSet::node2d &instances, int batch_size) override;
 
 };

diff --git a/include/thundersvm/model/svr.h b/include/thundersvm/model/svr.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_SVR_H
 #define THUNDERSVM_SVR_H
 
-#include "thundersvm/thundersvm-train.h"
+#include "thundersvm/thundersvm.h"
 #include "svmmodel.h"
 #include <map>
 

diff --git a/include/thundersvm/solver/csmosolver.h b/include/thundersvm/solver/csmosolver.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_CSMOSOLVER_H
 #define THUNDERSVM_CSMOSOLVER_H
 
-#include <thundersvm/thundersvm-train.h>
+#include <thundersvm/thundersvm.h>
 #include <thundersvm/kernelmatrix.h>
 
 class CSMOSolver {

diff --git a/include/thundersvm/svmparam.h b/include/thundersvm/svmparam.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_SVMPARAM_H
 #define THUNDERSVM_SVMPARAM_H
 
-#include "thundersvm-train.h"
+#include "thundersvm.h"
 
 struct SvmParam {
     SvmParam() {

diff --git a/include/thundersvm/syncdata.h b/include/thundersvm/syncdata.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_SYNCDATA_H
 #define THUNDERSVM_SYNCDATA_H
 
-#include "thundersvm-train.h"
+#include "thundersvm.h"
 #include "syncmem.h"
 
 template<typename T>

diff --git a/include/thundersvm/syncmem.h b/include/thundersvm/syncmem.h
@@ -5,7 +5,7 @@
 #ifndef THUNDERSVM_SYNCMEM_H
 #define THUNDERSVM_SYNCMEM_H
 
-#include "thundersvm-train.h"
+#include "thundersvm.h"
 
 class SyncMem {
 public:

diff --git a/include/thundersvm/thundersvm-train.h → include/thundersvm/thundersvm.h b/include/thundersvm/thundersvm-train.h → include/thundersvm/thundersvm.h
diff --git a/include/thundersvm/util/metric.h b/include/thundersvm/util/metric.h
@@ -8,6 +8,7 @@
 #include <thundersvm/thundersvm.h>
 
 class Metric {
+public:
     virtual string name() = 0;
 
     virtual real score(const vector<real> &predict_y, const vector<real> &ground_truth_y) = 0;

diff --git a/src/test/test_nusvc.cpp b/src/test/test_nusvc.cpp
@@ -42,17 +42,15 @@ class NuSVCTest : public ::testing::Test {
     }
 };
 
-//TEST_F(SVCTest, test_set) {
-//    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
-//                        "test_dataset.txt", DATASET_DIR
-//                        "test_dataset.txt", 100, 0.5), 0.98, 1e-5);
+//TEST_F(NuSVCTest, test_set) {
+//    load_dataset_and_train(DATASET_DIR "test_dataset.txt", DATASET_DIR "test_dataset.txt", 100, 0.5, 0.1);
 //}
 
-TEST_F(NuSVCTest, a9a) {
-    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
-                        "a9a", DATASET_DIR
-                        "a9a.t", 100, 0.5, 0.1), 0.826608, 1e-3);
-}
+//TEST_F(NuSVCTest, a9a) {
+//    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
+//                        "a9a", DATASET_DIR
+//                        "a9a.t", 100, 0.5, 0.1), 0.826608, 1e-3);
+//}
 //TEST_F(NuSVCTest, a1a1024) {
 //    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
 //                        "a1a1024", DATASET_DIR
@@ -62,7 +60,7 @@ TEST_F(NuSVCTest, a9a) {
 //TEST_F(NuSVCTest, mnist) {
 //    load_dataset_and_train(DATASET_DIR "mnist.scale", DATASET_DIR "mnist.scale.t", 10, 0.125, 0.1);
 //}
-//
+
 TEST_F(NuSVCTest, realsim) {
     load_dataset_and_train(DATASET_DIR "real-sim", DATASET_DIR "real-sim", 4, 0.5, 0.1);
 }

diff --git a/src/test/test_svc.cpp b/src/test/test_svc.cpp
@@ -44,18 +44,18 @@ class SVCTest : public ::testing::Test {
 //                        "test_dataset.txt", 100, 0.5), 0.98, 1e-5);
 //}
 
-TEST_F(SVCTest, a9a) {
-    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
-                        "a9a", DATASET_DIR
-                        "a9a.t", 100, 0.5), 0.826608, 1e-3);
-}
+//TEST_F(SVCTest, a9a) {
+//    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
+//                        "a9a", DATASET_DIR
+//                        "a9a.t", 100, 0.5), 0.826608, 1e-3);
+//}
 //
 //TEST_F(SVCTest, mnist) {
 //    load_dataset_and_train(DATASET_DIR "mnist.scale", DATASET_DIR "mnist.scale.t", 10, 0.125);
 //}
 //
-//TEST_F(SVCTest, realsim) {
-//    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
-//                        "real-sim", DATASET_DIR
-//                        "real-sim", 4, 0.5), 0.997276, 1e-3);
-//}
+TEST_F(SVCTest, realsim) {
+    EXPECT_NEAR(load_dataset_and_train(DATASET_DIR
+                        "real-sim", DATASET_DIR
+                        "real-sim", 4, 0.5), 0.997276, 1e-3);
+}
diff --git a/src/thundersvm/CMakeLists.txt b/src/thundersvm/CMakeLists.txt
@@ -2,7 +2,7 @@ set(CMAKE_ARCHIVE_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/lib)
 set(CMAKE_LIBRARY_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/lib)
 set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
 file(GLOB_RECURSE SRC *.c*)
-list(REMOVE_ITEM SRC ${CMAKE_CURRENT_SOURCE_DIR}/thundersvm-train.cpp)
+list(REMOVE_ITEM SRC ${CMAKE_CURRENT_SOURCE_DIR}/thundersvm-*.cpp)
 cuda_add_library(${PROJECT_NAME}_lib ${SRC})
 cuda_add_executable(${PROJECT_NAME}-train thundersvm-train.cpp ${COMMON_INCLUDES})
 cuda_add_executable(${PROJECT_NAME}-predict thundersvm-predict.cpp ${COMMON_INCLUDES})

diff --git a/src/thundersvm/cmdparser.cpp b/src/thundersvm/cmdparser.cpp
@@ -123,6 +123,7 @@ void CMDParser::parse_command_line(int argc, char **argv) {
                     break;
                 case 'u':
                     gpu_id = atoi(argv[i]);
+                    break;
                 default:
                     fprintf(stderr, "Unknown option: -%c\n", argv[i - 1][1]);
                     HelpInfo_svmtrain();
@@ -154,6 +155,9 @@ void CMDParser::parse_command_line(int argc, char **argv) {
                 case 'b':
 //                    predict_probability = atoi(argv[i]);
                     break;
+                case 'u':
+                    gpu_id = atoi(argv[i]);
+                    break;
                 default:
                     fprintf(stderr, "Unknown option: -%c\n", argv[i - 1][1]);
                     HelpInfo_svmpredict();
@@ -176,10 +180,12 @@ void CMDParser::parse_command_line(int argc, char **argv) {
         strcpy(svmpredict_input_file, argv[i]);
         strcpy(svmpredict_output_file, argv[i + 2]);
         strcpy(svmpredict_model_file_name, argv[i + 1]);
-    } else {
-        printf("Usage: thundersvm [options] training_set_file [model_file]\n"
-                       "or: thundersvm_predict [options] test_file model_file output_file\n"
-                       "or: thundersvm_scale [options] data_filename\n");
-        exit(0);
     }
+//    else {
+//
+//        printf("Usage: thundersvm [options] training_set_file [model_file]\n"
+//                       "or: thundersvm_predict [options] test_file model_file output_file\n"
+//                       "or: thundersvm_scale [options] data_filename\n");
+//        exit(0);
+//    }
 }
diff --git a/src/thundersvm/model/oneclass_svc.cu → src/thundersvm/model/oneclass_svc.cpp b/src/thundersvm/model/oneclass_svc.cu → src/thundersvm/model/oneclass_svc.cpp
@@ -6,8 +6,6 @@
 #include <thundersvm/model/oneclass_svc.h>
 #include <thundersvm/solver/csmosolver.h>
 
-using namespace std;
-
 void OneClassSVC::train(DataSet dataset, SvmParam param) {
     int n_instances = dataset.total_count();
     SyncData<real> alpha(n_instances);

diff --git a/src/thundersvm/model/svr.cpp b/src/thundersvm/model/svr.cpp
@@ -0,0 +1,39 @@
+//
+// Created by jiashuai on 17-10-5.
+//
+#include <iostream>
+#include <thundersvm/kernel/kernelmatrix_kernel.h>
+#include <thundersvm/solver/csmosolver.h>
+#include "thundersvm/model/svr.h"
+
+void SVR::train(DataSet dataset, SvmParam param) {
+    int n_instances = dataset.total_count();
+
+    //duplicate instances
+    DataSet::node2d instances_2(dataset.instances());
+    instances_2.insert(instances_2.end(), dataset.instances().begin(), dataset.instances().end());
+
+    KernelMatrix kernelMatrix(instances_2, param);
+
+    SyncData<real> f_val(n_instances * 2);
+    SyncData<int> y(n_instances * 2);
+
+    for (int i = 0; i < n_instances; ++i) {
+        f_val[i] = param.p - dataset.y()[i];
+        y[i] = +1;
+        f_val[i + n_instances] = -param.p - dataset.y()[i];
+        y[i + n_instances] = -1;
+    }
+
+    SyncData<real> alpha_2(n_instances * 2);
+    alpha_2.mem_set(0);
+    int ws_size = min(max2power(n_instances) * 2, 1024);
+    CSMOSolver solver;
+    solver.solve(kernelMatrix, y, alpha_2, rho, f_val, param.epsilon, param.C, param.C, ws_size);
+    SyncData<real> alpha(n_instances);
+    for (int i = 0; i < n_instances; ++i) {
+        alpha[i] = alpha_2[i] - alpha_2[i + n_instances];
+    }
+    record_model(alpha, y, dataset.instances(), param);
+}
+
diff --git a/src/thundersvm/model/svr.cu b/src/thundersvm/model/svr.cu