Loading [MathJax]/jax/output/HTML-CSS/jax.js
1 Star 2 Fork 0

dechin/CyFES

加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
克隆/下载
贡献代码
同步代码
取消
提示: 由于 Git 不支持空文件夾,创建文件夹后会生成空的 .keep 文件
Loading...
README
MIT

CyFES:GPU加速高性能数据透视工具

Hex.pm Hex.pm Hex.pm Hex.pm Hex.pm

介绍

这是一个使用Cython+CUDA+Python编写的高性能FES计算软件,可以用于加速数据透视的计算:

E(x)=kTlog(Nj=1eVjkTe(xnj)22σ2NΠi2πσi)

安装

pip安装

本项目可以直接使用pip进行安装:

$ python3 -m pip install cyfes --user --upgrade -i https://pypi.org/simple

源码安装

首先将本仓库clone到本地:

$ git clone https://gitee.com/dechin/cy-fes.git && cd cy-fes/

然后直接运行setup.py进行安装:

$ python3 -m pip install .

安装测试

在本仓库的test路径下存放了一个测试用例,用于测试cyfes是否安装成功。用户可以直接简单的运行:

$ python3 tests/test_path_fes.py 
[0.00902854 0.         0.09338432 0.02065182]
[0.03961432 0.         0.01514649 0.02259541]
[0.00129778 0.         0.02988457 0.0869869 ]
[0.01712827 0.01378975 0.         0.02229569]
[0.0114323  0.03356422 0.         0.0328276 ]

若输出为多个数组,则表示安装成功。也可以使用单元测试运行,但是这需要在本地先安装pytest

$ python3 -m pip install pytest

然后直接在仓库的根目录下运行:

$ py.test
============================ test session starts =============================
platform linux -- Python 3.7.5, pytest-7.4.4, pluggy-1.2.0
rootdir: /home/cy-fes
collected 5 items                                                            

tests/test_path_fes.py .....                                           [100%]

============================= 5 passed in 14.23s =============================

没有报错,则表示安装成功。

使用方法

在安装成功后,可以直接在Python脚本中调用:

import numpy as np
from cyfes import PathFES
np.random.seed(0)

def test_path_fes():
    atoms = 4
    cvs = 10000
    crd = np.random.random((atoms, 3))
    cv = np.random.random((cvs, 3))
    bw = np.random.random(3)
    bias = np.random.random(cvs)-1

    fes = np.asarray(PathFES(crd, cv, bw, bias))
    print (fes)

if __name__ == '__main__':
    test_path_fes()

还可以使用命令行模式:

$ python3 -m cyfes --help
usage: __main__.py [-h] [-i I] [-ic IC] [-ib IB] [-s S] [-e E] [-g G] [-o O]
                   [-no_bias NO_BIAS] [-f32 F32] [-sigma SIGMA]
                   [-device DEVICE]

optional arguments:
  -h, --help        show this help message and exit
  -i I              Set the input record file path.
  -ic IC            Set the cv index of input record file. Default: 0,1,2
  -ib IB            Set the bias index of input record file. Default: 3
  -s S              CV length. Default: None
  -e E              Edge length. Default: 1.0
  -g G              Grid numbers. Default: 10,10,10
  -o O              Set the output FES file path.
  -no_bias NO_BIAS  Do not use the bias from input file. Default: false
  -f32 F32          Use float32. Default: false
  -sigma SIGMA      Sigma value when calculating FES. Default: 0.3
  -device DEVICE    Set the device ids separated with commas. Default: 0

假如我们有一个三维的CV,那么最简单的运行方式为:

$ python3 -m cyfes -i /home/Data/xyz_bias.txt -o ./work_dir/z.cub

那么最后产生的文件内容为:

$ head -n 10 work_dir/z.cub
Generated by CyFES
Total	1000	grids
1	21.6622	19.8498	42.3652
10	6.40465	0	0
10	0	7.02147	0
10	0	0	6.33118
1	1.000000	53.6854	54.9571	74.0211
450	450	450	450	450	450	
450	450	450	450	450	450	
450	450	70.0855	70.848	450	450	

该cube格式的文件可以在支持的软件(如VMD)中进行可视化操作。

已知问题

  1. 使用numpy==1.22.2的版本中会出现ImportError: numpy.core.multiarray failed to import问题。解决方案:升级numpy版本:python3 -m pip install numpy --upgrade
  2. 执行python3 -m cyfes --help报错ModuleNotFoundError: No module named 'cyfes.wrapper',这是因为执行命令的目录下存在名为cyfes的文件夹,需要切换执行命令的位置。
  3. 使用cyfes出现Segmentation fault段错误问题,是因为找不到编译好的动态链接库文件,大概率是系统环境下权限不足,没有site路径的权限,可以使用如下脚本进行检查:
# check_dynamics.py
import os
import site
from pathlib import Path

site_path = Path(site.getsitepackages()[0])
site_file_path = site_path.parent.parent.parent / 'cyfes' / 'libcufes.so'
site_dynamics_path = str(site_file_path)

user_site_path = Path(site.USER_SITE)
user_file_path = user_site_path.parent.parent.parent / 'cyfes' / 'libcufes.so'
user_dynamics_path = str(user_file_path)

if not os.path.exists(site_dynamics_path) and not os.path.exists(user_dynamics_path):
    print ('Check dynamics complete, no libcufes.so file founded!')
else:
    print ('Installation of CyFES success!')

使用python3运行该脚本,即可判断动态链接库是否被正确安装。

MIT License Copyright (c) 2024 dechin Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

简介

CyFES是一款基于Python/Cython和CUDA混合开发的高性能自由能计算工具,面向用户开放Python API接口。 展开 收起
Python 等 3 种语言
MIT
取消

发行版 (2)

全部

贡献者

全部

近期动态

不能加载更多了
马建仓 AI 助手
尝试更多
代码解读
代码找茬
代码优化
Python
1
https://gitee.com/dechin/cy-fes.git
git@gitee.com:dechin/cy-fes.git
dechin
cy-fes
CyFES
master

搜索帮助