some signed unsigned warnings fixed

include/LightGBM/c_api.h

@@ -411,7 +411,7 @@ ColumnFunctionFromCSC(const void* col_ptr, int col_ptr_type, const int32_t* indi
   const void* data, int data_type, int64_t ncol_ptr, int64_t nelem);
 
 std::vector<double> 
-SampleFromOneColumn(const std::vector<std::pair<int, double>>& data, const std::vector<size_t>& indices);
+SampleFromOneColumn(const std::vector<std::pair<int, double>>& data, const std::vector<int>& indices);
 
 
 // exception handle and error msg

include/LightGBM/utils/random.h

@@ -52,12 +52,12 @@ public:
   * \param K
   * \return K Ordered sampled data from {0,1,...,N-1}
   */
-  inline std::vector<size_t> Sample(size_t N, size_t K) {
-    std::vector<size_t> ret;
+  inline std::vector<int> Sample(int N, int K) {
+    std::vector<int> ret;
     if (K > N || K < 0) {
       return ret;
     }
-    for (size_t i = 0; i < N; ++i) {
+    for (int i = 0; i < N; ++i) {
       double prob = (K - ret.size()) / static_cast<double>(N - i);
       if (NextDouble() < prob) {
         ret.push_back(i);

src/boosting/dart.hpp

@@ -88,7 +88,7 @@ private:
     // select dropping tree indexes based on drop_rate
     // if drop rate is too small, skip this step, drop one tree randomly
     if (drop_rate_ > kEpsilon) {
-      for (size_t i = 0; i < static_cast<size_t>(iter_); ++i) {
+      for (int i = 0; i < iter_; ++i) {
         if (random_for_drop_.NextDouble() < drop_rate_) {
           drop_index_.push_back(i);
         }
@@ -128,7 +128,7 @@ private:
     }
   }
   /*! \brief The indexes of dropping trees */
-  std::vector<size_t> drop_index_;
+  std::vector<int> drop_index_;
   /*! \brief Dropping rate */
   double drop_rate_;
   /*! \brief Random generator, used to select dropping trees */

src/c_api.cpp

@@ -199,7 +199,7 @@ DllExport int LGBM_CreateDatasetFromMat(const void* data,
   if (reference == nullptr) {
     // sample data first
     Random rand(config.io_config.data_random_seed);
-    const size_t sample_cnt = static_cast<size_t>(nrow < config.io_config.bin_construct_sample_cnt ? nrow : config.io_config.bin_construct_sample_cnt);
+    const int sample_cnt = static_cast<int>(nrow < config.io_config.bin_construct_sample_cnt ? nrow : config.io_config.bin_construct_sample_cnt);
     auto sample_indices = rand.Sample(nrow, sample_cnt);
     std::vector<std::vector<double>> sample_values(ncol);
     for (size_t i = 0; i < sample_indices.size(); ++i) {
@@ -251,7 +251,7 @@ DllExport int LGBM_CreateDatasetFromCSR(const void* indptr,
   if (reference == nullptr) {
     // sample data first
     Random rand(config.io_config.data_random_seed);
-    const size_t sample_cnt = static_cast<size_t>(nrow < config.io_config.bin_construct_sample_cnt ? nrow : config.io_config.bin_construct_sample_cnt);
+    const int sample_cnt = static_cast<int>(nrow < config.io_config.bin_construct_sample_cnt ? nrow : config.io_config.bin_construct_sample_cnt);
     auto sample_indices = rand.Sample(nrow, sample_cnt);
     std::vector<std::vector<double>> sample_values;
     for (size_t i = 0; i < sample_indices.size(); ++i) {
@@ -313,7 +313,7 @@ DllExport int LGBM_CreateDatasetFromCSC(const void* col_ptr,
     Log::Warning("Construct from CSC format is not efficient");
     // sample data first
     Random rand(config.io_config.data_random_seed);
-    const size_t sample_cnt = static_cast<size_t>(nrow < config.io_config.bin_construct_sample_cnt ? nrow : config.io_config.bin_construct_sample_cnt);
+    const int sample_cnt = static_cast<int>(nrow < config.io_config.bin_construct_sample_cnt ? nrow : config.io_config.bin_construct_sample_cnt);
     auto sample_indices = rand.Sample(nrow, sample_cnt);
     std::vector<std::vector<double>> sample_values(ncol_ptr - 1);
 #pragma omp parallel for schedule(guided)
@@ -762,7 +762,7 @@ ColumnFunctionFromCSC(const void* col_ptr, int col_ptr_type, const int32_t* indi
   throw std::runtime_error("unknown data type in ColumnFunctionFromCSC");
 }
 
-std::vector<double> SampleFromOneColumn(const std::vector<std::pair<int, double>>& data, const std::vector<size_t>& indices) {
+std::vector<double> SampleFromOneColumn(const std::vector<std::pair<int, double>>& data, const std::vector<int>& indices) {
   size_t j = 0;
   std::vector<double> ret;
   for (auto row_idx : indices) {

src/io/dataset_loader.cpp

@@ -512,11 +512,11 @@ std::vector<std::string> DatasetLoader::LoadTextDataToMemory(const char* filenam
 }
 
 std::vector<std::string> DatasetLoader::SampleTextDataFromMemory(const std::vector<std::string>& data) {
-  size_t sample_cnt = static_cast<size_t>(io_config_.bin_construct_sample_cnt);
-  if (sample_cnt > data.size()) {
-    sample_cnt = data.size();
+  int sample_cnt = io_config_.bin_construct_sample_cnt;
+  if (static_cast<size_t>(sample_cnt) > data.size()) {
+    sample_cnt = static_cast<int>(data.size());
   }
-  std::vector<size_t> sample_indices = random_.Sample(data.size(), sample_cnt);
+  auto sample_indices = random_.Sample(static_cast<int>(data.size()), sample_cnt);
   std::vector<std::string> out(sample_indices.size());
   for (size_t i = 0; i < sample_indices.size(); ++i) {
     const size_t idx = sample_indices[i];

src/metric/dcg_calculator.cpp

@@ -38,7 +38,7 @@ score_t DCGCalculator::CalMaxDCGAtK(data_size_t k, const float* label, data_size
   for (data_size_t i = 0; i < num_data; ++i) {
     ++label_cnt[static_cast<int>(label[i])];
   }
-  size_t top_label = label_gain_.size() - 1;
+  int top_label = static_cast<int>(label_gain_.size()) - 1;
 
   if (k > num_data) { k = num_data; }
   //  start from top label, and accumulate DCG
@@ -67,7 +67,7 @@ void DCGCalculator::CalMaxDCG(const std::vector<data_size_t>& ks,
   }
   score_t cur_result = 0.0f;
   data_size_t cur_left = 0;
-  size_t top_label = label_gain_.size() - 1;
+  int top_label = static_cast<int>(label_gain_.size()) - 1;
   // calculate k Max DCG by one pass
   for (size_t i = 0; i < ks.size(); ++i) {
     data_size_t cur_k = ks[i];

src/treelearner/serial_tree_learner.cpp

@@ -141,8 +141,8 @@ void SerialTreeLearner::BeforeTrain() {
     is_feature_used_[i] = false;
   }
   // Get used feature at current tree
-  size_t used_feature_cnt = static_cast<size_t>(num_features_*feature_fraction_);
-  std::vector<size_t> used_feature_indices = random_.Sample(num_features_, used_feature_cnt);
+  int used_feature_cnt = static_cast<int>(num_features_*feature_fraction_);
+  auto used_feature_indices = random_.Sample(num_features_, used_feature_cnt);
   for (auto idx : used_feature_indices) {
     is_feature_used_[idx] = true;
   }