Project: twip (GitHub Link)

twip-master
- CHANGES.rst
- .coveragerc
- requirements-test.txt
- .pre-commit-config.yaml
- test-requirements.txt
- manage.py
- run.sh
- README.rst
- setup.py
- data
  - geo.svg
  - tlds-from-iana.csv
  - geo_tweets.csv.gz
  - uri-schemes.xhtml.csv
  - tweet_vocab.csv.gz
- AUTHORS.rst
- setup.cfg
- .travis.yml
- README.md
- tests
  - travis_install.sh
  - test_util.py
  - conftest.py
- twip
  - features.py
  - util.py
  - tweets.py
  - wsgi.py
  - nlp.py
  - get_tweets.py
  - plot.py
  - urls.py
  - settings_template.py
  - train.py
  - __init__.py
  - constant.py
  - scripts
    - explore.py
    - skeleton.py
    - clean.py
    - hooks
      - pre-commit
    - plot_globe.py
    - cat_tweets.py
    - generate.py
    - __init__.py
  - settings.py
  - cleaner.py
  - futil.py
  - cosine_distance.py
  - wip
    - build_pycon_slides.py
- scripts
  - django-generate-secret-key.py
- tweetget
  - run_tweetget.sh
  - cron.py
  - single.py
  - __init__.py
  - README.md
  - settings.py
  - core.py
- requirements.txt
- index.html
- .gitignore
- docs
  - Makefile
  - slides
    - 0 - Introduction.md
    - build
    - 3 - Predict Impact.md
    - pandoc-template-for-revealjs.html
    - 2 - Got Features.md
    - pre-commit
    - X - Example RevealJS Markdown MathJax etc.md
    - 0 - Introduction.html
    - 4 - Advanced Techniques.md
    - 1 - Got Tweets.html
    - 1 - Got Tweets.md
    - .index.md
    - X - Example RevealJS Markdown MathJax etc.html
    - 4 - Advanced Techniques.html
    - 2 - Got Features.html
    - .index.html
    - 3 - Predict Impact.html
  - images
    - geo.svg
    - README.md
  - notebooks
    - 04 Data -- Natural Language.ipynb
    - 21 gensim -- Quanitfied Self.py
    - 06 Features -- Numeric.ipynb
    - 99 WIP -- Data Exploration.py
    - 11 Learning -- Regression.py
    - Shakespeare.ipynb
    - 10 Learning -- Predicting Impact.ipynb
    - 11 Learning -- Regression.ipynb
    - 04 Data -- Natural Language.py
    - 12b Learning -- More Features.py
    - 03 Data -- Getting Selective.ipynb
    - 02 Data -- Exploration.py
    - 07 Features -- Text.ipynb
    - setup_notebooks.py
    - shakescorpus.py
    - 21 gensim -- Quanitfied Self.ipynb
    - 08 Features -- TFIDF with Gensim.ipynb
    - jupyter_setup.py
    - 10 Learning -- Predicting Impact.py
    - constants.py
    - 01 Data -- Cleaning.ipynb
    - 06 Features -- Numeric.py
    - 09 Features -- LSI with Gensim.py
    - 05 Features -- Time.ipynb
    - __init__.py
    - 12 Advanced -- Word2Vec.py
    - 12b Learning -- More Features.ipynb
    - 01 Data -- Cleaning.py
    - 03 Data -- Getting Selective.py
    - 05 Features -- Time.py
    - 22 sentiment -- Quantified Self.ipynb
    - 02 Data -- Exploration.ipynb
    - summarization_tutorial.ipynb
    - 12 Advanced -- Word2Vec.ipynb
    - 08 Features -- TFIDF with Gensim.py
    - 07 Features -- Text.py
  - changes.rst
  - license.rst
  - authors.rst
  - index.rst
  - conf.py
  - references
    - formal-language.md
    - 2015-10-21-Biomedical-Literature-NLP.md
    - requirements-modulegraph-dependency-tree.md
    - random-notes.md
    - tweet-values.md
    - 2015-12-08-Idea--Baysean-TFIDF.md
    - tweet-timing.md
    - feedback-from-Rob.md
    - Making-Connections-with-Natural-Language-Processing.md
    - accepted-proposal.txt
    - gensim-ideas.md
    - README.md
    - proposal.md
    - tweet-impact.md
    - applications.md
    - ideas.md
    - 2015-12-08-B-Machine.md
  - _static
    - .gitignore
  - repl-sessions
    - pandas_str_accessor.py
    - tfidf-with-gensim.txt
- LICENSE.txt
- requirements
  - old-gensim-requirements.txt
- tox.ini

twip

Tweet Impact Predictor

DESCRIPTION

A natural language processing pipeline for predicting the impact (reach and popularity) of a tweet. Built as part of the PyCon 2016 Natural Language Processing tutorial and workshop.

Don't install the latest version from PyPi if you're working through the tutorial yourself! Tagged version numbers will correspond to sections of the tutorial and handout material so you can maintain pace even if you miss a step along the way. Plus it'll be easier to set up your API keys if you clone the repository.

GETTING STARTED

Rather than installing this module from the cheese shop, clone it to your laptop.

git clone git@github.com:totalgood/twip.git
# optional:
# mkvirtualenv twip
pip install -e twip
cd twip/docs/notebooks
ipython notebook

That way you can edit the source code. Even better, make your own fork so you can easily issue pull requests. Obviously it needs a lot of help.

GOT TWEETS?

To use the tweetget app, you also need a Twiter API key.

If you don't already have one, sign up to get a twitter user account like @yournewusername:

twitter.com/signup

And we'll be happy to be your first followers, just tweet us at:

Hobson Lane: [@hobsonlane]((https://twitter.com/hobsonlane)
Jeremy Robin: [@robusican]((https://twitter.com/robustican)
Dan Fellin: @eupharis

Once you have a user account, sign into it, then create a new twitter app with an API_KEY:

apps.twitter.com/app/new

Copy and paste the Consumer API Key and Consumer API Secret into the indicated places in the file called settings_template.py but don't save it there. Instead save the file as a new file named settings_secret.py. This file is .gitignored during pushes. Do a git status to make sure you didn't accidentally save your secret KEYs in the template file or misname your settings_secret.py file. If you see that any tracked/added files have changes then you need to undo them before you do a commit and push to your fork of twip.

Alternatively, check out the settings_secret.py file for the environment variables you can set to hold these secret values.

CREDITS

Hobson Lane Data Scientist @ Talentpair
Dan Fellin Engineer @ Talentpair
Jeremy Robin CTO @ Talentpair
Talentpair
Rob Ludwick: Proposal editting
PyScaffold: Python packages done right
gensim: The Star of the Show