[docs] fixed miscellaneous typos in comments and documentation (#2496)

* fixed miscellaneous typos in documentation

* fix typo introduced in typo-fixing PR

include/LightGBM/tree_learner.h

@@ -46,7 +46,7 @@ class TreeLearner {
   virtual void ResetConfig(const Config* config) = 0;
 
   /*!
-  * \brief training tree model on dataset 
+  * \brief training tree model on dataset
   * \param gradients The first order gradients
   * \param hessians The second order gradients
   * \param is_constant_hessian True if all hessians share the same value
@@ -56,7 +56,7 @@ class TreeLearner {
                       const Json& forced_split_json) = 0;
 
   /*!
-  * \brief use a existing tree to fit the new gradients and hessians.
+  * \brief use an existing tree to fit the new gradients and hessians.
   */
   virtual Tree* FitByExistingTree(const Tree* old_tree, const score_t* gradients, const score_t* hessians) const = 0;
 

include/LightGBM/utils/array_args.h

@@ -14,7 +14,7 @@
 namespace LightGBM {
 
 /*!
-* \brief Contains some operation for a array, e.g. ArgMax, TopK.
+* \brief Contains some operation for an array, e.g. ArgMax, TopK.
 */
 template<typename VAL_T>
 class ArrayArgs {

include/LightGBM/utils/common.h

@@ -595,7 +595,7 @@ inline static int64_t Pow2RoundUp(int64_t x) {
 }
 
 /*!
- * \brief Do inplace softmax transformaton on p_rec
+ * \brief Do inplace softmax transformation on p_rec
  * \param p_rec The input/output vector of the values.
  */
 inline static void Softmax(std::vector<double>* p_rec) {

src/network/linkers.h

@@ -30,7 +30,7 @@
 namespace LightGBM {
 
 /*!
-* \brief An network basic communication warpper.
+* \brief A network basic communication warpper.
 * Will warp low level communication methods, e.g. mpi, socket and so on.
 * This class will wrap all linkers to other machines if needs
 */
@@ -52,7 +52,7 @@ class Linkers {
   * \brief Recv data, blocking
   * \param rank Which rank will send data to local machine
   * \param data Pointer of receive data
-  * \prama len Recv size, will block until recive len size of data
+  * \param len Recv size, will block until recive len size of data
   */
   inline void Recv(int rank, char* data, int len) const;
 
@@ -62,7 +62,7 @@ class Linkers {
   * \brief Send data, blocking
   * \param rank Which rank local machine will send to
   * \param data Pointer of send data
-  * \prama len Send size
+  * \param len Send size
   */
   inline void Send(int rank, char* data, int len) const;
 
@@ -71,10 +71,10 @@ class Linkers {
   * \brief Send and Recv at same time, blocking
   * \param send_rank
   * \param send_data
-  * \prama send_len
+  * \param send_len
   * \param recv_rank
   * \param recv_data
-  * \prama recv_len
+  * \param recv_len
   */
   inline void SendRecv(int send_rank, char* send_data, int send_len,
                        int recv_rank, char* recv_data, int recv_len);

src/treelearner/gpu_tree_learner.h

@@ -110,7 +110,7 @@ class GPUTreeLearner: public SerialTreeLearner {
 
   /*!
    * \brief Returns OpenCL kernel build log when compiled with option opts
-   * \param opts OpenCL build options 
+   * \param opts OpenCL build options
    * \return OpenCL build log
   */
   std::string GetBuildLog(const std::string &opts);
@@ -120,7 +120,7 @@ class GPUTreeLearner: public SerialTreeLearner {
   */
   void SetupKernelArguments();
 
-  /*! 
+  /*!
    * \brief Compute GPU feature histogram for the current leaf.
    *        Indices, gradients and hessians have been copied to the device.
    * \param leaf_num_data Number of data on current leaf
@@ -136,7 +136,7 @@ class GPUTreeLearner: public SerialTreeLearner {
   void WaitAndGetHistograms(HistogramBinEntry* histograms);
 
   /*!
-   * \brief Construct GPU histogram asynchronously. 
+   * \brief Construct GPU histogram asynchronously.
    *        Interface is similar to Dataset::ConstructHistograms().
    * \param is_feature_used A predicate vector for enabling each feature
    * \param data_indices Array of data example IDs to be included in histogram, will be copied to GPU.
@@ -144,9 +144,9 @@ class GPUTreeLearner: public SerialTreeLearner {
    * \param num_data Number of data examples to be included in histogram
    * \param gradients Array of gradients for all examples.
    * \param hessians Array of hessians for all examples.
-   * \param ordered_gradients Ordered gradients will be generated and copied to GPU when gradients is not nullptr, 
+   * \param ordered_gradients Ordered gradients will be generated and copied to GPU when gradients is not nullptr,
    *                     Set gradients to nullptr to skip copy to GPU.
-   * \param ordered_hessians Ordered hessians will be generated and copied to GPU when hessians is not nullptr, 
+   * \param ordered_hessians Ordered hessians will be generated and copied to GPU when hessians is not nullptr,
    *                     Set hessians to nullptr to skip copy to GPU.
    * \return true if GPU kernel is launched, false if GPU is not used
   */
@@ -189,13 +189,13 @@ class GPUTreeLearner: public SerialTreeLearner {
   /*! \brief Currently used kernel name */
   std::string kernel_name_;
 
-  /*! \brief a array of histogram kernels with different number
+  /*! \brief an array of histogram kernels with different number
      of workgroups per feature */
   std::vector<boost::compute::kernel> histogram_kernels_;
-  /*! \brief a array of histogram kernels with different number
+  /*! \brief an array of histogram kernels with different number
      of workgroups per feature, with all features enabled to avoid branches */
   std::vector<boost::compute::kernel> histogram_allfeats_kernels_;
-  /*! \brief a array of histogram kernels with different number
+  /*! \brief an array of histogram kernels with different number
      of workgroups per feature, and processing the whole dataset */
   std::vector<boost::compute::kernel> histogram_fulldata_kernels_;
   /*! \brief total number of feature-groups */
@@ -210,7 +210,7 @@ class GPUTreeLearner: public SerialTreeLearner {
   /*! \brief total number of dense feature-group tuples on GPU.
    * Each feature tuple is 4-byte (4 features if each feature takes a byte) */
   int num_dense_feature4_;
-  /*! \brief Max number of bins of training data, used to determine 
+  /*! \brief Max number of bins of training data, used to determine
    * which GPU kernel to use */
   int max_num_bin_;
   /*! \brief Used GPU kernel bin size (64, 256) */