Python implementation of Gale and Sampson's (1995/2001) "Simple Good Turing" algorithm.
Python
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
README
sgt.py

README

This module provides an implementation of Gale and Sampson's (1995/2001) "Simple
Good Turing" algorithm. The main function is simpleGoodTuringProbs(), which
takes a dictionary of species counts and returns the estimated population
frequencies of the species, as estimated by the Simple Good Turing method. To
use this module, you must have scipy and numpy installed.

Also included is a function that uses pylab and matplotlib to draw a useful
scatterplot for comparing the empirical frequencies against the Simple Good
Turing estimates.

Depends on reasonably recent versions of scipy and numpy.

Version 0.3: June 21, 2011
    First github version.

Version 0.2: November 12, 2009. 
    Added __version__ string.
    Added check for 0 counts.
    Don't pollute namespace with "import *".
    Added loglog keyword argument to plotFreqVsGoodTuring().
Version 0.1: November 11, 2009.

REFERENCES:
    William Gale and Geoffrey Sampson. 1995. Good-Turing frequency estimation
    without tears. Journal of Quantitative Linguistics, vol. 2, pp. 217--37.
    
    See also the corrected reprint of same on Sampson's web site.