Proxy setting for R

2019-01-02 20:00发布

I am facing problem while conecting R with internet in my office. May be this due to LAN settings. I tried the almost all possible ways I come across in the web (see below) but still in vain.

  • Method1: Invoking R using --internet2

  • Method2: Invoking R by setting ~/Rgui.exe http_proxy=http:/999.99.99.99:8080/ http_proxy_user=ask

  • Method3: Setting Setinternet2=TRUE

  • Method4:

    curl <- getCurlHandle()
    curlSetOpt(.opts = list(proxy = '999.99.99.99:8080'), curl = curl)
    Res <- getURL('http://www.cricinfo.com', curl = curl)
    

In above all methods I can able to load packages directly from CRAN also able to download files using download.file command

But using getURL(RCurl), readHTMLTable(XML), htmlTreeParse(XML) commands I am unable to extract web data. I am getting ~<HEAD>\n<TITLE>Access Denied</TITLE>\n</HEAD>~ error.

How to set LAN proxy settings for XML package in R?

标签: r
11条回答
唯独是你
2楼-- · 2019-01-02 20:16

Inspired by all the responses related on the internet, finally I've found the solution to correctly configure the Proxy for R and Rstudio.

There are several steps to follow, perhaps some of the steps are useless, but the combination works!

  1. Add environment variables http_proxy and https_proxy with proxy details.

    variable name: http_proxy
    variable value: https://user_id:password@your_proxy:your_port/
    
    variable name: https_proxy
    variable value: https:// user_id:password@your_proxy:your_port
    
  2. If you start R from a desktop icon, you can add the --internet flag to the target line (right click -> Properties)

    e.g."C:\Program Files\R\R-2.8.1\bin\Rgui.exe" --internet2

  3. For RStudio just you have to do this:

    Firstly, open RStudio like always, select from the top menu:

    Tools-Global Options-Packages

    Uncheck the option: Use Internet Explorer library/proxy for HTTP

  4. Find the file (.Renviron) in your computer, most probably you would find it here: C:\Users\your user name\Documents.

    Note that: if it does not exist you can create it just by writing this command in R:

    file.edit('~/.Renviron')
    

    Then add these six lines to the initials of the file:

    options(internet.info = 0)
    
    http_proxy = https:// user_id:password@your_proxy:your_port
    
    http_proxy_user = user_id:password
    
    https_proxy = https:// user_id:password0@your_proxy:your_port
    
    https_proxy_user = user_id:password
    
    ftp_proxy = user_id:password@your_proxy:your_port
    
  5. Restart R. Type the following commands in R to assure that the configuration above works well:

    Sys.getenv("http_proxy")
    
    Sys.getenv("http_proxy_user")
    
    Sys.getenv("https_proxy")
    
    Sys.getenv("https_proxy_user")
    
    Sys.getenv("ftp_proxy")
    
  6. Now you can install the packages as you want by using the command like:

    install.packages("mlr",method="libcurl")
    

    It's important to add method="libcurl", otherwise it won't work.

查看更多
梦醉为红颜
3楼-- · 2019-01-02 20:16

Simplest way to get everything working in RStudio under Windows 10:

Open up Internet Explorer, select Internet Options:

enter image description here


Open editor for Environment variables:

enter image description here


Add a variable HTTP_PROXY in form:

HTTP_PROXY=http://username:password@localhost:port/

Example:

HTTP_PROXY=http://John:JohnPassword@localhost:8080/    

enter image description here


RStudio should work:

enter image description here

查看更多
何处买醉
4楼-- · 2019-01-02 20:19

Tried all of these and also the solutions using netsh, winhttp etc. Geek On Acid's answer helped me download packages from the server but none of these solutions worked for using the package I wanted to run (twitteR package).

The best solution is to use a software that let's you configure system-wide proxy.

FreeCap (free) and Proxifier (trial) worked perfectly for me at my company.

Please note that you need to remove proxy settings from your browser and any other apps that you have configured to use proxy as these tools provide system-wide proxy for all network traffic from your computer.

查看更多
有味是清欢
5楼-- · 2019-01-02 20:22

This post pertains to R proxy issues on *nix. You should know that R has many libraries/methods to fetch data over internet.

For 'curl', 'libcurl', 'wget' etc, just do the following:

  1. Open a terminal. Type the following command:

    sudo gedit /etc/R/Renviron.site
    
  2. Enter the following lines:

    http_proxy='http://username:password@abc.com:port/'
    https_proxy='https://username:password@xyz.com:port/'
    

    Replace username, password, abc.com, xyz.com and port with these settings specific to your network.

  3. Quit R and launch again.

This should solve your problem with 'libcurl' and 'curl' method. However, I have not tried it with 'httr'. One way to do that with 'httr' only for that session is as follows:

library(httr)
set_config(use_proxy(url="abc.com",port=8080, username="username", password="password"))

You need to substitute settings specific to your n/w in relevant fields.

查看更多
只若初见
6楼-- · 2019-01-02 20:23

If you start R from a desktop icon, you can add the --internet flag to the target line (right click -> Properties) e.g.

"C:\Program Files\R\R-2.8.1\bin\Rgui.exe" --internet2 
查看更多
登录 后发表回答