- 积分
- 3601
- 贡献
-
- 精华
- 在线时间
- 小时
- 注册时间
- 2012-9-8
- 最后登录
- 1970-1-1
|
登录后查看更多精彩内容~
您需要 登录 才可以下载或查看,没有帐号?立即注册
x
直译为公式翻译器的科学计算语言Fortran虽几经兴衰,但仍奋力与时俱进。它与近年崛起的统计语言-R可谓各有千秋。在专职统计功能方面,R语言似乎胜出一筹、风头强劲。于是乎,热心人士发起了强强联合计划,RFortran行动应运而生,并设立了专门网站(http://www.rfortran.org),现简介如下:
RFortran is an open source project with the following aim:
To provide an easy to use Fortran software library that enables Fortran programs to transfer data and commands to and from R.
Since R is an open source and ever-growing repository of routines for statistical, numerical and graphical analysis, RFortran provides numerous benefits to users, including:
Graphical functions: Produce high quality graphs from Fortran, automate reporting and expedite analysis of large datasets,
Numerical and statistical algorithms: R packages can be used to quickly extend the functionality of Fortran programs.
Requirements- Windows XP (Service Pack 3) or Windows 7 32/64bit (untested on Windows Vista)
- Administrator Rights for initial installation (to install and registor COM server)
- R (>=2.12.0) (32 bit only - see issue 90), download here
- Intel Visual Fortran (IVF) >=v10.0 and Microsoft Visual Studio (VS) >=2005
OR - Compaq Visual Fortran (CVF) v6.6
Note: IVF (>=v11.0) and Visual Studio 2008 (VS2008) are needed for example project/solutions, download here If you use Visual Studio 2005 (VS2005) and/or IVF 10.X you can follow the instructions hereto change the VS2008 solution and IVF11 project files to work with VS2005 and IVF10
Download and Installation
- Download latest version of RFortran installer here.
- If any errors occur during installation see manual installation below
- If installation reported no errors please see Getting Started for how to test RFortran.
Manual InstallationInstaller automates most of theses steps. If there is a problem follow these manual instructions to check installation is complete (internet connection is required)
Note: If you connect to the internet via a proxy, see R Windows FAQ for details on setting proxies for R
- Install/Re-install R and ensure 'Save version number in Registry' is selected.
- Install statconn (download here) and install R packages, rcom and rscproxy, and register COM connection
- Done semi-automatically by running R script RFortran\setup\install_rcom.r
OR - Manually download and install packages (rcom and rscproxy) and statconn
- After installing rcom package in R, type command comRegisterRegistry() this should return NULL - this registers the COM connection
- Add library(rcom) to Rpath\etc\Rprofile.site file.
- Ensures rcom library is loaded every time R is started
- (Optional) To install additional R packages required by RFortran's extended functionality run the R script RFortran\setup\installpackages_RFortran.r
- Requires a large internet download and can take some time.
- Install the source code to the RFortran folder on your PC
- General User: Use the RFortran installer and choose "RFortran source"
- Developers: Checkout source code from svn repository
- Checkout instructions here - note this installs the latest development version from the trunk
- It is recommended to download the latest release version located on a branch
- In folder RFortran\Settings\ copy file RFortran.set_DEFAULT and rename to RFortran.set
- Change last line of RFortran.set file to the path to Rgui.exe (e.g. C:\Program Files\R\R-12.2.0\bin\i386\)
- Create an environment variable (Control Panel|System|Advanced|Environmental Variables) RFORTRAN_PATH and set its value to the install folder of RFortran
Upgrading to a new version of RFortranGeneral Advice for Users- In case of issues with the latest version it is recommended to keep back-up copies of:
- Uninstall previous versions of RFortran (from Start Menu)
- Install new version of RFortran to same folder as the previous version
- maintains filepaths to static library/source code for your projects that use RFortran
- Note if you have not changed versions of R you don't need to install RCOM server again, only the RFortran source
- Test the new version using Tutorial Examples
- Test the new version using your own projects
- If there any problems:
General Advice for Developers
- Since v2.0.0 RFortran makes use of branches for official release versions.
- This means official release (more stable) versions are now located on a branch folder of the repository while the trunk folder is used for development of latest features
- If you want to use a stable version it is recommended to work from the branch version folder
- Further information is given in the Branching Approach document
- When a new version of RFortran is released you can do a fresh checkout or switch your working copy to checkout from the latest branch version folder
- After performing this checkout of the new version you will need to copy and rename RFortran.set_DEFAULT to RFortran.set and change RGuiPath variable to match your local R installation
Upgrading from RFortran prior to v1.0RFortran v1.0.X is a major update from <v1.0, with numerous source code changes, therefore it is recommended to take the following action
- First install RFortran to a dummy or test folder and make sure everything is working by using the tutorial examples
- Make a back-up copy of your previous version of RFortran
- Install the new version to the same folder
- Test the code using your own projects, see Version History for any backwards compatibility issues
Upgrading to a new version of RIf you want RFortran to work with your latest version of R, follow these steps:
- Install the new R version
- Update the RCOM server to link with the latest version of R by either
- Re-running the RFortran installer and only install the component: RCOM server
OR - Follow steps 1-4 and 7 in the manual installation
continued by topmad
官网的视频我已经搞下来了...放到了迅雷快传http://kuai.xunlei.com/d/7TORBgKTBQCSd5hRef7...
网页版正在上传中 稍后奉上
|
评分
-
查看全部评分
|