---
output:
  md_document:
    variant: gfm
---

# Accessing Twitter Text

### Kenneth Benoit
### 24 April 2017

### Overview

In this section we will cover the following methods for mining and analyzing text from the twitter API:

* Creating a new application and getting access to the REST and streaming APIs
* Using the REST and streaming APIs from within R

### Accessing the twitter APIs

To access the REST and streaming APIs, you will need to create a twitter application, and generate authentication credentials associated with this application. To do this you will first need to have a twitter account. You will also need to install at least the following R packages: twitteR, 


```{r, eval = FALSE}
install.packages(c('twitteR', 'streamR', 'RCurl', 'ROAuth', 'httr'))
```

To register a twitter application and get your consumer keys:

1. Go to <https://apps.twitter.com> in a web browser.
2. Click on 'create new app'.
3. Give your app a unique name, a description, any relevant web address, and agree to the terms and conditions. Set the callback URL to http://127.0.0.1:1410. You might have to add a cellphone number your twitter account.
4. Go to the keys and access section of the app page, and copy your consumer key and consumer secret to the code below.
5. (optional): For actions requiring write permissions, generate an access token and access secret .

```{r, eval = FALSE}
require(twitteR)
require(streamR)
require(ROAuth)

consumerKey <- 'your key here'
consumerSecret <- 'your secret here'

# Try this first, to use twitteR
setup_twitter_oauth(consumerKey, consumerSecret)
results <- searchTwitter('#DemDebate')
df <- as.data.frame(t(sapply(results, as.data.frame)))
```

Then try these instructions, to use streamR:
<https://github.com/pablobarbera/streamR#installation-and-authentication>