fix bug in filter bin

include/LightGBM/utils/common.h

@@ -425,8 +425,8 @@ inline static double ApproximateHessianWithGaussian(const double y, const double
 }
 
 template <typename T>
-inline static T** Vector2Ptr(std::vector<std::vector<T>>& data) {
-  T** ptr = new T*[data.size()];
+inline static std::vector<T*> Vector2Ptr(std::vector<std::vector<T>>& data) {
+  std::vector<T*> ptr(data.size());
   for (size_t i = 0; i < data.size(); ++i) {
     ptr[i] = data[i].data();
   }

src/c_api.cpp

@@ -423,8 +423,8 @@ LIGHTGBM_C_EXPORT int LGBM_DatasetCreateFromMat(const void* data,
       }
     }
     DatasetLoader loader(io_config, nullptr, 1, nullptr);
-    ret.reset(loader.CostructFromSampleData(Common::Vector2Ptr<double>(sample_values), 
-                                            Common::Vector2Ptr<int>(sample_idx), 
+    ret.reset(loader.CostructFromSampleData(Common::Vector2Ptr<double>(sample_values).data(), 
+                                            Common::Vector2Ptr<int>(sample_idx).data(), 
                                             static_cast<int>(sample_values.size()),
                                             Common::VectorSize<double>(sample_values).data(),
                                             sample_cnt, nrow));
@@ -487,8 +487,8 @@ LIGHTGBM_C_EXPORT int LGBM_DatasetCreateFromCSR(const void* indptr,
     }
     CHECK(num_col >= static_cast<int>(sample_values.size()));
     DatasetLoader loader(io_config, nullptr, 1, nullptr);
-    ret.reset(loader.CostructFromSampleData(Common::Vector2Ptr<double>(sample_values),
-                                            Common::Vector2Ptr<int>(sample_idx),
+    ret.reset(loader.CostructFromSampleData(Common::Vector2Ptr<double>(sample_values).data(),
+                                            Common::Vector2Ptr<int>(sample_idx).data(),
                                             static_cast<int>(sample_values.size()),
                                             Common::VectorSize<double>(sample_values).data(),
                                             sample_cnt, nrow));
@@ -546,8 +546,8 @@ LIGHTGBM_C_EXPORT int LGBM_DatasetCreateFromCSC(const void* col_ptr,
       }
     }
     DatasetLoader loader(io_config, nullptr, 1, nullptr);
-    ret.reset(loader.CostructFromSampleData(Common::Vector2Ptr<double>(sample_values),
-                                            Common::Vector2Ptr<int>(sample_idx),
+    ret.reset(loader.CostructFromSampleData(Common::Vector2Ptr<double>(sample_values).data(),
+                                            Common::Vector2Ptr<int>(sample_idx).data(),
                                             static_cast<int>(sample_values.size()),
                                             Common::VectorSize<double>(sample_values).data(),
                                             sample_cnt, nrow));

src/io/bin.cpp

@@ -49,18 +49,14 @@ bool NeedFilter(std::vector<int>& cnt_in_bin, int total_cnt, int filter_cnt, Bin
     int sum_left = 0;
     for (size_t i = 0; i < cnt_in_bin.size() - 1; ++i) {
       sum_left += cnt_in_bin[i];
-      if (sum_left >= filter_cnt) {
-        return false;
-      } else if (total_cnt - sum_left >= filter_cnt) {
+      if (sum_left >= filter_cnt && total_cnt - sum_left >= filter_cnt) {
         return false;
       }
     }
   } else {
     for (size_t i = 0; i < cnt_in_bin.size() - 1; ++i) {
       int sum_left = cnt_in_bin[i];
-      if (sum_left >= filter_cnt) {
-        return false;
-      } else if (total_cnt - sum_left >= filter_cnt) {
+      if (sum_left >= filter_cnt && total_cnt - sum_left >= filter_cnt) {
         return false;
       }
     }

src/io/config.cpp

@@ -141,8 +141,8 @@ void OverallConfig::CheckParamConflict() {
   bool objective_type_multiclass = (objective_type == std::string("multiclass"));
   int num_class_check = boosting_config.num_class;
   if (objective_type_multiclass) {
-    if (num_class_check <= 2) {
-      Log::Fatal("Number of classes should be specified and greater than 2 for multiclass training");
+    if (num_class_check <= 1) {
+      Log::Fatal("Number of classes should be specified and greater than 1 for multiclass training");
     }
   } else {
     if (task_type == TaskType::kTrain && num_class_check != 1) {

src/io/dataset_loader.cpp

@@ -487,7 +487,9 @@ Dataset* DatasetLoader::CostructFromSampleData(double** sample_values,
       feature_names_.push_back(str_buf.str());
     }
   }
-  const data_size_t filter_cnt = static_cast<data_size_t>(static_cast<double>(0.95 * io_config_.min_data_in_leaf) / num_data * num_col);
+
+  const data_size_t filter_cnt = static_cast<data_size_t>(
+    static_cast<double>(io_config_.min_data_in_leaf * total_sample_size) / num_data);
 
   #pragma omp parallel for schedule(guided)
   for (int i = 0; i < num_col; ++i) {
@@ -701,7 +703,8 @@ void DatasetLoader::ConstructBinMappersFromTextData(int rank, int num_machines,
   }
   dataset->feature_names_ = feature_names_;
   std::vector<std::unique_ptr<BinMapper>> bin_mappers(sample_values.size());
-  const data_size_t filter_cnt = static_cast<data_size_t>(static_cast<double>(0.95 * io_config_.min_data_in_leaf) / dataset->num_data_ * sample_values.size());
+  const data_size_t filter_cnt = static_cast<data_size_t>(
+    static_cast<double>(io_config_.min_data_in_leaf* sample_values.size()) / dataset->num_data_);
 
   // start find bins
   if (num_machines == 1) {
@@ -815,7 +818,7 @@ void DatasetLoader::ConstructBinMappersFromTextData(int rank, int num_machines,
     }
   }
   sample_values.clear();
-  dataset->Construct(bin_mappers, Common::Vector2Ptr<int>(sample_indices),
+  dataset->Construct(bin_mappers, Common::Vector2Ptr<int>(sample_indices).data(),
                      Common::VectorSize<int>(sample_indices).data(), sample_data.size(), io_config_);
 }