When loading a binary file, take feature penalty and monotone constraints from...

When loading a binary file, take feature penalty and monotone constraints from config if given there. (#1881)

* When loading a binary file, take feature penalty from config if given there.

* When loading a binary file, take feature penalty from config if given there.

* Fix crash when num_features != num_total_features and feature_contri is given.

* Apply the same logic to monotone_types_.

* Fix indentation

src/io/dataset_loader.cpp

@@ -370,10 +370,22 @@ Dataset* DatasetLoader::LoadFromBinFile(const char* data_filename, const char* b
   }
   mem_ptr += sizeof(int) * (dataset->num_groups_);
 
-  const int8_t* tmp_ptr_monotone_type = reinterpret_cast<const int8_t*>(mem_ptr);
-  dataset->monotone_types_.clear();
-  for (int i = 0; i < dataset->num_features_; ++i) {
-    dataset->monotone_types_.push_back(tmp_ptr_monotone_type[i]);
+  if(!config_.monotone_constraints.empty()) {
+    CHECK(dataset->num_total_features_ == config_.monotone_constraints.size());
+    dataset->monotone_types_.resize(dataset->num_features_);
+    for(int i = 0; i < dataset->num_total_features_; ++i){
+      int inner_fidx = dataset->InnerFeatureIndex(i);
+      if(inner_fidx >= 0) {
+        dataset->monotone_types_[inner_fidx] = config_.monotone_constraints[i];
+      }
+    }
+  }
+  else {
+    const int8_t* tmp_ptr_monotone_type = reinterpret_cast<const int8_t*>(mem_ptr);
+    dataset->monotone_types_.clear();
+    for (int i = 0; i < dataset->num_features_; ++i) {
+      dataset->monotone_types_.push_back(tmp_ptr_monotone_type[i]);
+    }
   }
   mem_ptr += sizeof(int8_t) * (dataset->num_features_);
 
@@ -381,10 +393,22 @@ Dataset* DatasetLoader::LoadFromBinFile(const char* data_filename, const char* b
     dataset->monotone_types_.clear();
   }
 
-  const double* tmp_ptr_feature_penalty = reinterpret_cast<const double*>(mem_ptr);
-  dataset->feature_penalty_.clear();
-  for (int i = 0; i < dataset->num_features_; ++i) {
-    dataset->feature_penalty_.push_back(tmp_ptr_feature_penalty[i]);
+  if(!config_.feature_contri.empty()) {
+    CHECK(dataset->num_total_features_ == config_.feature_contri.size());
+    dataset->feature_penalty_.resize(dataset->num_features_);
+    for(int i = 0; i < dataset->num_total_features_; ++i){
+      int inner_fidx = dataset->InnerFeatureIndex(i);
+      if(inner_fidx >= 0) {
+        dataset->feature_penalty_[inner_fidx] = config_.feature_contri[i];
+      }
+    }
+  }
+  else {
+    const double* tmp_ptr_feature_penalty = reinterpret_cast<const double*>(mem_ptr);
+    dataset->feature_penalty_.clear();
+    for (int i = 0; i < dataset->num_features_; ++i) {
+      dataset->feature_penalty_.push_back(tmp_ptr_feature_penalty[i]);
+    }
   }
   mem_ptr += sizeof(double) * (dataset->num_features_);