Linux环境下搭建R计算平台

      R语言已经广泛地应用于数据分析与文本挖掘,Facebook与Google的挖掘工具都是R语言,它较之于S-Plus与SAS具有开源等优势,具能布置于云平台下进行计算工作。与Matlab等一样,R语言也是从Linux下向Windows发展,但由于许多的R语言程序包只能在Linux下才能用,今天我特意在个人的电脑上搭建了VMware虚拟机下的Fedora14系统,进行实验。

Step1:VMware下搭建Fedora

    安装VMware Station后在虚拟机的平台上安装Fedora14-i686.iso包,过程不再赘述!

    具体的安装参考可见:http://blog.csdn.net/ppzlyg/article/details/4957792

安装R有两种方法,一种是取rpm包,另一种方法则是通过安装R-2-15-1.tar.gz文件。

Step 2:安装R语言环境

方法一:

    1、下载R语言包(http://cran.r-project.org/bin/linux/redhat/fedora11/i386/),把目录下所有的.rpm文件都下载到一个文件夹中,如/home/libing/R下。我用的是Fedora,并且是32位机,所以选择这个。如果是64位机,则应该选择X86-64文件夹下的.rpm文件。

    2、安装之前的准备工作:

         R语言在安装之前,有许多的依赖包要用,所以这个步骤是用yum安装这些依赖包

         a.  安装gcc g++ gfortran编译器:

                   yum install gcc

                   yum install gcc-c++

                   yum install gcc-gfortran

         b.  yum install pcre-devel

               yum install tcl-devel

               yum install zlib-devel

               yum install bzip2-devel

               yum install libX11-devel

               yum install readline-devel    #否则报”–with-readline=yes (default) and headers/libs are not available”错误
               yum install libXt-devel          #否则报”configure: error: –with-x=yes (default) and X11 headers/libs are not available”错误

               yum install tk-devel

               yum install tetex-latex

         c. 上面准备工作已经做好了,可以开始安装R语言了:

               [root@rac1 linux_rpm]# rpm -ivh  libRmath-2.10.0-2.fc11.i586.rpm    

                [root@rac1 linux_rpm]# rpm -ivh  libRmath-devel-2.10.0-2.fc11.i586.rpm   //这两步是安装llibRmath包,直接安装,没有依赖包

               [root@rac1 linux_rpm]# rpm -Uvh R-core-2.10.0-2.fc11.i586.rpm     //安装R-core时可能会发生错误,其Error信息如下:

                 

                          warning: R-core-2.10.0-2.fc11.i586.rpm:Header V3 DSA/SHA1 Signature, key ID 97d3544e: NOKEY

                          error: Failed dependencies:

                          libtcl8.4.so()(64bit) is needed by R-core-2.10.0-2.fc11.i586

                          libtk8.4.so()(64bit) is needed by R-core-2.10.0-2.fc11.i586

                          perl(File::Copy::Recursive) is needed by R-core-2.10.0-2.fc11.i586

                         --这里安装报错,说要8.4 的tcl 包,实际我们已经通过yum安装了8.5的,所以这里我们可以强制安装。

                         [root@rac1 linux_rpm]# rpm -qa|grep tcl

                          tcl-devel-8.5.7-6.fc11.i586

                          tcl-8.5.7-6.fc11.i586

                         强制安装R-core:

                   [root@rac1 linux_rpm]# rpm -Uvh R-core-2.10.0-2.fc11.i586.rpm --force --nodeps

                            warning: R-core-2.10.0-2.fc11.i586.rpm:Header V3 DSA/SHA1 Signature, key ID 97d3544e: NOKEY

                            Preparing... ###########################################[100%]

                            1:R-core########################################### [100%]

                   [root@rac1 linux_rpm]# rpm -ivh R-devel-2.10.0-2.el5.x86_64.rpm   //安装R-devel

                   [root@rac1 linux_rpm]# rpm -ivh R-2.10.0-2.el5.x86_64.rpm              //最后安装R

          d.测试

               [root@rac1 ~]# xhost +

               access control disabled, clients canconnect from any host

               [root@rac1 ~]# R

               R version 2.10.0 (2009-10-26)

               Copyright (C) 2009 The R Foundation forStatistical Computing

               ISBN 3-900051-07-0

               R is free software and comes withABSOLUTELY NO WARRANTY.

               You are welcome to redistribute it undercertain conditions.

               Type 'license()' or 'licence()' fordistribution details.

               Natural language support but running in an English locale

               R is a collaborative project with manycontributors.

               Type 'demo()' for some demos, 'help()' foron-line help, or

               'help.start()' for an HTML browserinterface to help.

               Type 'q()' to quit R.

                >(接下来的操作与R在Windows相同)

 方法二:

           下载R-2-15.1.tar.gz文件,解压

                            tar -zvxf  R-2-15.1.tar.gz   //生成了一个R-2-15-1的文件夹

                             cd R-2-15-1

                             ./configure

                              make

                              make install

              通过这种方法,比较简单地就能安装R.   在以后的安装新包的过程中,使用指令 install.packages(“RODBC”) install.packages(“tseries”)即可矣!

Step 3:实验操作,选择一些只针对于Linux平台才能发挥作用的一些R包进行操作实验(未完待续!!!)



              

             

        



發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章