dP""b8    db    88      dP""b8 88   88 88        db    888888 888888 
dP   `"   dPYb   88     dP   `" 88   88 88       dPYb     88   88__   
Yb       dP__Yb  88  .o Yb      Y8   8P 88  .o  dP__Yb    88   88""   
 YboodP dP""""Yb 88ood8  YboodP `YbodP' 88ood8 dP""""Yb   88   888888

Some simple math we use to do journalism.

• Documentation: http://latimes-calculate.rtfd.org
• Issues: https://github.com/datadesk/latimes-calculate/issues
• Packaging: https://pypi.python.org/pypi/latimes-calculate
• Testing: https://travis-ci.org/datadesk/latimes-calculate
• Coverage: https://coveralls.io/r/datadesk/latimes-calculate

Features

• Descriptive statistics like mean, median, percentile, mode, range, standard deviation
• Comparison statistics like percentage change, per-capita, per square mile, percentiles, deciles and rankings
• Geospatial stats like mean center and standard deviation distance
• A small dab of more complicated hoohah like Pearson’s R
• A grabbag of utilities for a diversity index, Benford’s Law, ages, margin of victory, date rates, making break points, generating random points and other things

Dependencies

For most functions, nothing. GeoDjango and its dependencies are required for a small number of the geospatial functions, though the rest of the module will work if it is not installed.

Getting started

Install from PyPI

$ pip install latimes-calculate

Experiment in Python shell

>>> import calculate
>>> calculate.percentage_change(100, 150)
50.0