#1

Hi, i need help with my research. I am using standard twitter API. But I need tweets in the past. From about September 2019 to today. I state that I don’t know how to use python, I’m using R. My university could finance the premium API. which one should i do? I also need different data because I follow different #. In addition, I would also need research through gelocation. Thanks in advance.

#2

I would not recommend R for gathering data. The existing libraries don’t work with Premium very well (you will end up using more calls than necessary, and it will cost more because of it)

I recomend Twarc Collect Twitter Data with Twarc! · Learn Twarc! (it’s python but you don’t need to know python, just enough to install it and then use it as a command line tool)

It supports making standard and premium calls. Once you have the data in json, you can import it into R where you’re more used to performing whatever analysis you need.

#3

Thank you, so you confirm that also with python is not possible to go back in the past more tha 30 days?
I was looking my packages. I have sandbox and is looking in the past like the premium with 30 days. How can I go more in the past?

#4

or maybe someone knows If still exist some collection of data regarding automation, digital transformation or robots? Also this could be usefull. Becouse I need the text of the tweets and the geolocation for run a sentiment analysis and a regional analysis.

#5

Twarc can use Any of these search APIs, except “enterprise”: Overview | Docs | Twitter Developer Platform

So the standard search can only go back 7 days: GitHub - DocNow/twarc: A command line tool (and Python library) for archiving Twitter JSON

30day Premium can go back 30 days, and fullarchive Premium can go back to 2006 to the very first tweets. GitHub - DocNow/twarc: A command line tool (and Python library) for archiving Twitter JSON

To get older than 30 days you would use Premium fullarchive search. But i’d still recommend trying experiments with the 30day Premium endpoint in sandbox before paying for it - to make sure your code for gathering data works ok.

There are a few existing collections on https://catalog.docnow.io/ that might be useful. And a whole bunch of sentiment analysis ones in existance, but maybe not on the topic you want - they might still be useful though. Geo data is harder to come by, as very few tweets have geo information.