https://github.com/emfomy/isvd
Integrated Singular Value Decomposition (iSVD)
https://github.com/emfomy/isvd
distributed-algorithm gpu parallel-algorithm randomized-algorithm singular-value-decomposition
Last synced: about 1 month ago
JSON representation
Integrated Singular Value Decomposition (iSVD)
- Host: GitHub
- URL: https://github.com/emfomy/isvd
- Owner: emfomy
- License: mit
- Created: 2017-08-03T07:09:30.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2019-09-11T07:33:11.000Z (over 5 years ago)
- Last Synced: 2025-03-12T16:06:46.732Z (2 months ago)
- Topics: distributed-algorithm, gpu, parallel-algorithm, randomized-algorithm, singular-value-decomposition
- Language: C
- Size: 3.14 MB
- Stars: 2
- Watchers: 2
- Forks: 1
- Open Issues: 18
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
Awesome Lists containing this project
README
# Integrated Singular Value Decomposition (iSVD)
# Information
This library is a C implementation of the Integrated Singular Value Decomposition (iSVD), which is an parallel algorithm for computing low-rank approximate singular value decomposition of large size matrices.
## Git
* https://github.com/emfomy/isvd
## Documentation
* Download **isvd-#.#.#-docs.zip** from https://github.com/emfomy/isvd/releases/latest
* May also build it using **make doc** (see below) on your own computer.
## Author
* Mu Yang <>
## Tutorial
* \ref tutorial_main
# Requirements
* [CMake](https://cmake.org) 2.8.11+ (CMake 3.0+ recommended).
* C/C++ compiler with C99/C++98 standard support ([GCC](https://gcc.gnu.org) 4.4+ recommended).
* [BLAS](http://www.netlib.org/blas) & [LAPACK](http://www.netlib.org/lapack) library (Used for multi-core parallelization).
* MPI Library with 2.2+ standard support ([MPICH](http://www.mpich.org) 1.2+ or [OpenMPI](https://www.open-mpi.org) 1.7+).
## Optional
* [Intel® Math Kernel Library](https://software.intel.com/en-us/intel-mkl) (**Recommended** for better performance, used for BLAS & LAPACK).
* [MAGMA](http://icl.cs.utk.edu/magma/) 2.0+ (Used for BLAS & LAPACK with GPU support).
* [OpenMP](http://openmp.org) Library (**Recommended** for better performance, used for multi-thread parallelization).
* [Google Test Library](https://github.com/google/googletest) (Used for code testing).
* [Doxygen](http://www.stack.nl/~dimitri/doxygen/) (Used for documentation).
# Installation
Please use the following commands to create Makefiles
```
cd
mkdir build
cd build
cmake ..
```
## Options
Use the following command to set options
```
ccmake ..
```
The following table are the main options
| Option | Detail | Note |
|------------------------|------------------------------------|--------------------------------|
| `ISVD_BLAS` | the selection BLAS/LAPACK library. | Options: `BLAS/MKL` |
| `ISVD_BUILD_DEMO` | build demo codes. | |
| `ISVD_BUILD_DOC` | build documentation. | |
| `ISVD_BUILD_LIB` | build libraries. | |
| `ISVD_BUILD_TEST` | build unit tests. | |
| `ISVD_INDEX_TYPE` | the selection index type. | Options: `32/64` |
| `ISVD_OMP` | the selection OpenMP library. | Options: `OFF/GOMP/IOMP` |
| `ISVD_USE_GPU` | enable GPU support. | |
| `ISVD_TEST_VERBOSE` | enable verbose unit tests. | |
| `GTEST_ROOT` | the root path of Google Test. | Require `ISVD_BUILD_TEST` |
| `INTEL_ROOT` | the root path of Intel libraries. | Require `ISVD_BLAS = MKL` |
| `MKL_ROOT` | the root path of Intel MKL. | Require `ISVD_BLAS = MKL` |
| `MPI_PROCS` | the number of MPI processes. | Only used in demo codes. |
| `OMP_THRDS` | the number of OpenMP threads. | Only used in demo/check codes. |
## Makefile
The following table are the main Makefile rules
| Command | Detail | Options |
|----------------|--------------------------------|--------------------------------|
| `make all` | build all libraries | |
| `make install` | install package | |
| `make check` | build and run unit tests | Require `ISVD_BUILD_TEST` |
| `make doc` | build documentation | Require `ISVD_BUILD_DOC` |
| `make help` | display make-rules | |
## Test installation
* Set `ISVD_BUILD_TEST` using **ccmake**. (Also recommended to unset `ISVD_TEST_VERBOSE` if GPU is enabled).
* Run **make check**
* Known issue: **RealSingle_WenYinIntegration.Test.#** / **s_integrate_wen_yin_#** fail the test.
# Usage
* Define `ISVD_USE_ILP64` before include `isvd.h` to use 64-bit integer.
* All 64bit libraries and executables are named with suffix `_64`.
* The header files are located in `build/include`
* The libraries are located in `build/lib`
## Libraries
Please link exactly one library for all categories.
* Correct
* `gcc test.c -I -L -Wl,-rpath -lisvd_core -lisvd_la_mkl_iomp -lisvd_gpu_magma`
* `gcc test.c -I -L -Wl,-rpath -lisvd_core -lisvd_la_blas -lisvd_gpu_none`
* Wrong
* `gcc test.c -I -L -Wl,-rpath -lisvd_core`
* (Some category of libraries are not linked)
* `gcc test.c -I -L -Wl,-rpath -lisvd_core -lisvd_la_blas -lisvd_gpu_magma -lisvd_gpu_none`
* (Some category of libraries are linked multiple times)
| Name | Category | Detail | Options |
|------------------------|----------------|----------------------------------------|------------------------------------|
| `isvd_core` | Core | Core routines | |
| `isvd_la_blas` | Linear Algebra | Sequential Plain BLAS | `ISVD_BLAS=BLAS`, `ISVD_OMP=NO` |
| `isvd_la_mkl` | Linear Algebra | Sequential Intel MKL | `ISVD_BLAS=MKL`, `ISVD_OMP=NO` |
| `isvd_la_mkl_gomp` | Linear Algebra | Parallel Intel MKL using GNU OpenMP | `ISVD_BLAS=MKL`, `ISVD_OMP=GOMP` |
| `isvd_la_mkl_iomp` | Linear Algebra | Parallel Intel MKL using Intel OpenMP | `ISVD_BLAS=MKL`, `ISVD_OMP=IOMP` |
| `isvd_gpu_none` | GPU | No GPU | `ISVD_USE_GPU=NO` |
| `isvd_gpu_magma` | GPU | MAGMA GPU | `ISVD_USE_GPU=YES` |
# Q&A
## How to set CMake options?
* Use `ccmake ..` or `ccmake ` in the `build` folder.
* Press `` on the option you want to change, change it, and press `` again.
* After changing the options, press `` to configure, and press `` to generate Makefiles if configured successfully.
* To quit without saving, press ``.
## Why isn't Intel MKL found?
* Source `mklvars.sh` in the `bin` folder of your Intel MKL to set the environment variables.
## Why isn't Google Test found?
* Set `GTEST_ROOT` to a folder containing `include` and `lib` of Google Test.
## How to enable multithread support?
* Set `ISVD_OMP` with `ccmake` before building libraries.
* Make sure your LAPACK&BLAS / Intel MKL uses supports multithreading.
## How to use 64-bit integer?
* Set `ISVD_USE_ILP64` with `ccmake` before building libraries; or
* Add `-DISVD_USE_ILP64` to compile flag.
* Make sure your LAPACK&BLAS / Intel MKL uses 64bit integer.
## Error "make[2]: *** No rule to make target 'tmp/......" occurs. How to solve it?
* Use `cmake .` to and recompile again.
* Note that the `tmp` folder will be deleted after `make clean`.
# Reference
* [Ting-Li Chen, Dawei D. Chang, Su-Yun Huang, Hung Chen, Chienyao Lin, Weichung Wang, “Integrating Multiple Random Sketches for Singular Value Decomposition”](https://arxiv.org/abs/1608.08285)
* [Mu Yang, “Highly Scalable Parallelism of Integrated Randomized Singular Value Decomposition with Big Data Applications”](http://doi.org/10.6342/NTU201702960)
# License {#Readme_License}
* [MIT License](LICENSE.md)
\include LICENSE.md