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

• 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

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.

Install from PyPI

$ pip install latimes-calculate

Experiment in Python shell

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