Skip to main content

Connecting to Twitter with R and RStudio

Social Media and Analytics is always an interesting combination. Twitter has an API that you can so you can connect to  using R and RStudio (and a whole bunch of other languages too) to analyse twitter data.

In this post we will simply look at how we go about connecting to twitter with R and Studio

Assuming you R and R Studio downloaded and installed

You need a Developer Account

Well actually, first up you'll need a twitter account. I'll assume that you have one. If you don't head over to twitter and signup. 

With your twitter account setup the first things first you need to do is create an application in twitter get  access to the Twitter API. This is a simple process. To do this you need to visit https://apps.twitter.com signing with your twitter account if you haven't done so already . When you have signed in, click on the create APP button.  

You will see a form called Create an Application. You will need to provide some application details:
  • Name - give your app a name
  • Description - describe the application
  • URL - this can be anything. I used my blog URL
  • Call back URL - I didn't provide this
The first three in the list are mandatory the callback url in optional.

Create a twitter application


Install TwitteR

This is great package for working with Twitter and the Twitter API. You need to install from the CRAN repository. Allow all  the dependant packages to be installed too.

Install twitteR Application
With twitteR installed the next you need to do is connect using R. To do this you need generate some access key from the twitter application that we created above. Go to the application and click on the the Key and Access Token

Access token and secret

API key and secret

The next step is to open up a new session in R Studio. The following code loads the TwitteR and DevTools packages and authenticates using your application keys

#Load libraries
library(devtools)
library(twitteR)

#Authenticate with twitter
api_key <- font="" your_api_key_here="">
api_secret <- font="" your_api_secret_here="">
access_token <- font="" your_access_token_here="">
access_token_secret <- font="" your_access_token_here="">
setup_twitter_oauth(api_key,api_secret,access_token,access_token_secret)

Then You can use the twitteR package to work with Twitter. The following updates your status. We will look at what else we can do with this package in a later post.

updateStatus("Another Morning of #R...#RStudio and twitteR")


Here's my tweet from RStudio

My tweet from Studio

R and twitteR can be a really useful combination in analysing your twitter feed. We'll look at a practical use for this in a later post.








Comments

  1. Hi!
    Thanks for the explanation! I've followed the steps and succeded to post in Twitter from RStudio! I'm waiting for the next post.
    just a detail: for an absolute newbie like me, the syntax of the variable assignation would be clearer if listed like:
    api_key <- "XXXXXX"
    Thanks again and greetings from Argentina
    Jair

    ReplyDelete

Post a Comment

Popular posts from this blog

SQL Server 2012 and Virtual Service Accounts

This post is written by David Postlethwaite
If you are using SQL Server 2012 you will probably have noticed that the default account for the SQL services has changed from that used in previous versions. With SQL 2005 and 2008 the default account for SQL service and SQL Agent service was “NT Authority\System”. This is one the built in accounts on a Windows machine, managed by the machine and selectable from a dedicated dropdown list

The Network Service account was introduced in Windows 2003 as an alternative to using the LocalSystem account, which has full local system privileges on the local machine, a major security concern.
The Network Service has limited local privileges easing these security concerns but when many services on a machine use the Network Service account it becomes harder to track which service is actually accessing resources and performing actions, because all the services are using the one Network Service account.
Also, this account, by default, has sysadmin per…

Always Encrypted

By David Postlethwaite

Always Encrypted is new features in SQL Server 2016 and it is also available in Azure SQL Database. Here you can encrypt columns in a table with a master key and a certificate so that they will appear as encrypted strings to those who don’t have the required certificate installed on their pc.
Once the certificate is installed on the computer then the unencrypted data can then be seen as normal.

The data passes from database to your application as the encrypted value, only the application with the correct certificate can unencrypt the data so it is secure across the wire. This will go some way to resolving the concern of people worried about putting their sensitive data on a shared server in the cloud such as Microsoft Azure and accessing the data across the Internet.

At the time of writing Always Encrypted is only supported with ADO.NET 4.6, JDBC 6.0 and ODBC 13.1 but expect other driver to become available.

The calling application (including SSMS) must also hav…

New in SQL Server 2017: Graph Databases

David has recorded and published a video of his presentation on SQL Server Graph Database. In his video which you can watch below, David provides an excellent introduction into SQL Server 2017 Graph Databases. In his presentation he looks at Tennis results at tournaments for  his favourite player "The Fed"  Rodger Federer.

David  shows how to set up graph database and work with them in SQL Server 2017.

Graph Database is not new. Other vendors have had graph database capabilities for some time so Microsoft are quite late to the market. In David presentation it appears that Microsoft have done a reasonable job of implementing some of the graph database features but he does point some of the limitations of the Microsoft product too and suggests that it is not ready for production yet but Microsoft seem serious about this feature.

Please watch the video and feel free to leave a comment or feedback - David is delivering a version of this talk on Graph databases in SQL Saturday Ka…