Require

• g++ 9.* (support C++17 and C++ parallelism Extension) or clang++ 8.0.1
• Eigen3 3.3.7 with intel mkl and intel tbb
  Eigen official download [http://eigen.tuxfamily.org/index.php?title=Main_Page#Download]
  To use intel mkl as eigen's backend, refer to [https://eigen.tuxfamily.org/dox/TopicUsingIntelMKL.html]
  To use intel tbb as libstdc++ parallelism backend, refer to [https://github.com/intel/tbb]
• cmake 3.12
• Boost
• python2.7 development headers and some packages
  On Ubuntu, install python requirement using the following command.

$ sudo apt-get install python-matplotlib python-numpy python-tk python2.7-dev

• GSL(GNU Scientific Library)
  On Ubuntu, install gsl using the following command.

$ sudo apt-get install libgsl-dev

Build

$ git clone 
$ cd lsh-cpp
$ git submodule update --init --recursive
$ mkdir build
$ cd build
$ cmake -DCMAKE_BUILD_TYPE=Release ..
(use cmake -DCMAKE_BUILD_TYPE=Debug .. to build in debug mode)
$ make

Run

$ cd build
$ ./lsh_cpp_test # run lsh test case
$ ./lsh_benchmark # run lsh benchmark

TODO

• LSH_CPP implementation

  • LSH benchmark test
  • LSH Forest impl
  • LSH ensemble impl
  • Weight MinHash impl
  • Lean MinHash impl (减少内存使用/压缩数据及参数/内存缓冲池)
  • HyperLog / HyperLog++ impl

• LSH_CPP Large-scale data support

  • Map reduce compute
  • data persistence (redis storage)
  • data save/read/remove/split/merge/... feature

• Other

  • 使用C++17 parallelism TS 加速部分for-loop/sort/reduce/find算法.
  • 加入对 xxhash , parallel_hash_map/set 的可选控制, 当用户编译不指定这些三方库时改用std:: hash和std:: unordered_map.
  • Python interface export using pybind11
  • WeightMinHash 实现中需要 concept 的部分待重构(比如WeightMinHash 的update()参数需要SetElementType存在weight()和value()接口)