python-string-similarity/similarity/cosine.py

# Copyright (c) 2018 luozhouyang
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in all
# copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.

import math

from .shingle_based import ShingleBased
from .string_distance import NormalizedStringDistance
from .string_similarity import NormalizedStringSimilarity

import re
_SPACE_PATTERN = re.compile("\\s+")


class Cosine(ShingleBased, NormalizedStringDistance,
             NormalizedStringSimilarity):

    def __init__(self, k):
        super().__init__(k)

    def distance(self, s0, s1):
        return 1.0 - self.similarity(s0, s1)

    def similarity(self, s0, s1):
        if s0 is None:
            raise TypeError("Argument s0 is NoneType.")
        if s1 is None:
            raise TypeError("Argument s1 is NoneType.")
        if s0 == s1:
            return 1.0
            
        s0 = _SPACE_PATTERN.sub("", s0)
        s1 = _SPACE_PATTERN.sub("", s1)
        
        if len(s0) < self.get_k() or len(s1) < self.get_k():
            return 0.0
        profile0 = self.get_profile(s0)
        profile1 = self.get_profile(s1)
        return self._dot_product(profile0, profile1) / (
                self._norm(profile0) * self._norm(profile1))

    def similarity_profiles(self, profile0, profile1):
        return self._dot_product(profile0, profile1) / (
                self._norm(profile0) * self._norm(profile1))

    @staticmethod
    def _dot_product(profile0, profile1):
        small = profile1
        large = profile0
        if len(profile0) < len(profile1):
            small = profile0
            large = profile1
        agg = 0.0
        for k, v in small.items():
            i = large.get(k)
            if not i:
                continue
            agg += 1.0 * v * i
        return agg

    @staticmethod
    def _norm(profile):
        agg = 0.0
        for k, v in profile.items():
            agg += 1.0 * v * v
        return math.sqrt(agg)