fix bug for get feature name from freed dataset

src/boosting/gbdt.cpp

@@ -107,6 +107,8 @@ void GBDT::ResetTrainingData(const BoostingConfig* config, const Dataset* train_
     max_feature_idx_ = train_data->num_total_features() - 1;
     // get label index
     label_idx_ = train_data->label_idx();
+    // get feature names
+    feature_names_ = train_data->feature_names();
   }
 
   if ((train_data_ != train_data && train_data != nullptr)
@@ -472,14 +474,8 @@ std::string GBDT::DumpModel(int num_iteration) const {
   str_buf << "\"max_feature_idx\":" << max_feature_idx_ << "," << std::endl;
   str_buf << "\"sigmoid\":" << sigmoid_ << "," << std::endl;
 
-  // output feature names
-  auto feature_names = std::ref(feature_names_);
-  if (train_data_ != nullptr) {
-    feature_names = std::ref(train_data_->feature_names());
-  }
-
   str_buf << "\"feature_names\":[\"" 
-     << Common::Join(feature_names.get(), "\",\"") << "\"]," 
+     << Common::Join(feature_names_, "\",\"") << "\"]," 
      << std::endl;
 
   str_buf << "\"tree_info\":[";
@@ -521,12 +517,8 @@ void GBDT::SaveModelToFile(int num_iteration, const char* filename) const {
   }
   // output sigmoid parameter
   output_file << "sigmoid=" << sigmoid_ << std::endl;
-  // output feature names
-  auto feature_names = std::ref(feature_names_);
-  if (train_data_ != nullptr) {
-    feature_names = std::ref(train_data_->feature_names());
-  }
-  output_file << "feature_names=" << Common::Join(feature_names.get(), " ") << std::endl;
+
+  output_file << "feature_names=" << Common::Join(feature_names_, " ") << std::endl;
 
   output_file << std::endl;
   int num_used_model = static_cast<int>(models_.size());
@@ -619,10 +611,7 @@ void GBDT::LoadModelFromString(const std::string& model_str) {
 }
 
 std::vector<std::pair<size_t, std::string>> GBDT::FeatureImportance() const {
-  auto feature_names = std::ref(feature_names_);
-  if (train_data_ != nullptr) {
-    feature_names = std::ref(train_data_->feature_names());
-  }
+
   std::vector<size_t> feature_importances(max_feature_idx_ + 1, 0);
     for (size_t iter = 0; iter < models_.size(); ++iter) {
         for (int split_idx = 0; split_idx < models_[iter]->num_leaves() - 1; ++split_idx) {
@@ -633,7 +622,7 @@ std::vector<std::pair<size_t, std::string>> GBDT::FeatureImportance() const {
     std::vector<std::pair<size_t, std::string>> pairs;
     for (size_t i = 0; i < feature_importances.size(); ++i) {
       if (feature_importances[i] > 0) {
-        pairs.emplace_back(feature_importances[i], feature_names.get().at(i));
+        pairs.emplace_back(feature_importances[i], feature_names_[i]);
       }
     }
     // sort the importance

src/treelearner/parallel_tree_learner.h

@@ -22,7 +22,7 @@ class FeatureParallelTreeLearner: public SerialTreeLearner {
 public:
   explicit FeatureParallelTreeLearner(const TreeConfig* tree_config);
   ~FeatureParallelTreeLearner();
-  virtual void Init(const Dataset* train_data);
+  void Init(const Dataset* train_data) override;
 
 protected:
   void BeforeTrain() override;