ToxTweet- Tools for Toxicovigliance on Twitter
Python files for Toxicovigilance on Twitter
What is it?
A collection of Python files, also listed on PyPI, as part of an ongoing effort to use Twitter for toxicology research.
How do I use it?
The easiest thing is to copy all of the files to someplace in your PYTHONPATH, like ~/bin.
RUN PYTHON MAIN.PY
This module creates a twitteryQuery.py object for all drugs listed in the formulary. It can save the results in JSON or XLS.
What other Python Plugins do I need?
-Numpy/SciPy -XLRD XLWT -python-twitter (use pip install twitter not easy_install twitter) -matplotlib -simplejson -networkx
What do the files besides main.py do?
—>twitterQuery.py: main.py creates an instance of twitterQuery for each drug in the “formulary” that you are … querying Twitter for.
- —>parser.py loads the JSON format results from querying twitter and formats them for NLP analysis by:
- Writing a reduplicated record of the tweets to an XLS file for human rating. (Change filename, which indicates the location of the JSON dump of the twitter query, as needed.)
- Making a stemmed corpus to use for semantic analysis, contextual analysis or as input to a classifier. (Change the savename as needed.)
—>loadTweets.py is a helper function to calculate the Fleisch-Kincaid grade level of a tweet.
—> medgraph.py draws a degree-rank plot with the social graph as an inset. ATTRIBUTION: This is a modified version of a file from the matplotlib examples gallery.