Coverage for nltk.metrics.spearman : 53%

# Natural Language Toolkit: Spearman Rank Correlation 

# 

# Copyright (C) 2001-2012 NLTK Project 

# Author: Joel Nothman <jnothman@student.usyd.edu.au> 

# URL: <http://nltk.org> 

# For license information, see LICENSE.TXT 

""" 

Tools for comparing ranked lists. 

""" 

def _rank_dists(ranks1, ranks2): 

    """Finds the difference between the values in ranks1 and ranks2 for keys 

    present in both dicts. If the arguments are not dicts, they are converted 

    from (key, rank) sequences. 

    """ 

    ranks1 = dict(ranks1) 

    ranks2 = dict(ranks2) 

    for k in ranks1: 

        try: 

            yield k, ranks1[k] - ranks2[k] 

        except KeyError: 

            pass 

def spearman_correlation(ranks1, ranks2): 

    """Returns the Spearman correlation coefficient for two rankings, which 

    should be dicts or sequences of (key, rank). The coefficient ranges from 

    -1.0 (ranks are opposite) to 1.0 (ranks are identical), and is only 

    calculated for keys in both rankings (for meaningful results, remove keys 

    present in only one list before ranking).""" 

    n = 0 

    res = 0 

    for k, d in _rank_dists(ranks1, ranks2): 

        res += d * d 

        n += 1 

    try: 

        return 1 - (6 * float(res) / (n * (n*n - 1))) 

    except ZeroDivisionError: 

        # Result is undefined if only one item is ranked 

        return 0.0 

def ranks_from_sequence(seq): 

    """Given a sequence, yields each element with an increasing rank, suitable 

    for use as an argument to ``spearman_correlation``. 

    """ 

    return ((k, i) for i, k in enumerate(seq)) 

def ranks_from_scores(scores, rank_gap=1e-15): 

    """Given a sequence of (key, score) tuples, yields each key with an 

    increasing rank, tying with previous key's rank if the difference between 

    their scores is less than rank_gap. Suitable for use as an argument to 

    ``spearman_correlation``. 

    """ 

    prev_score = None 

    rank = 0 

    for i, (key, score) in enumerate(scores): 

        try: 

            if abs(score - prev_score) > rank_gap: 

                rank = i 

        except TypeError: 

            pass 

        yield key, rank 

        prev_score = score