750words-analysis

An assortment of Python code to look at what you've written on 750words.com.

Download your writing

The 750 Words export files can be grabbed automatically if you have the Python packages requests, lxml, and pyquery installed, using this script:

$ python download_750words.py --help

Statistics

Text analysis code is in m750.py.

>>> from m750 import *
>>> text, entries = read_local_750words() # or text, entries = download_750words()

Statistics are calculated from the entries list via three classes::

>>> s750 = stats_750(entries)       # Streaks, etc.
>>> all = stats(text)               # Text statistics on all your writing
>>> estats = entrystats(entries)    # Text statistics per entry

Use print to explore the results of these classes, and take a look at their docstrings and methods. The graphing methods require matplotlib.

An example of a word cloud plotted using pytagcloud

You can also see some example IPython Notebooks:

• Analyse 750 Words content.ipynb
• word clouds via pytagcloud.ipynb
• convert to daily markdown files.ipynb
• how to do stop words list.ipynb
• Moby

The stats on the 750words.com website are -- and will probably continue to be -- far better. I made this framework mainly so I could look at how whatever measures I can find change over time, something the 750words.com site doesn't quite do at the moment (Feb 2013).

I've included hyphenation and parts-of-speech dictionaries from the Moby Project and some functions to load these in moby.py. The stats classes will use this data if you use use_moby=True. They're not included in the repository but you can download this 26 MB file and unzip it into a "Moby" folder here.

Viewing your writing and metadata

There is also a script which shows your entries and metadata on a single page through a local HTTP server. Just run:

$ python view750.py

Your entries

Your metadata