Method union

join/LSH/datasketch/minhash.py:270–290

Create a MinHash which is the union of the MinHash objects passed as arguments. Args: *mhs: The MinHash objects to be united. The argument list length is variable, but must be at least 2. Returns: datasketch.MinHash: A new union MinHash.

(cls, *mhs)

    @classmethod
    def union(cls, *mhs):
        '''Create a MinHash which is the union of the MinHash objects passed as arguments.

        Args:
            *mhs: The MinHash objects to be united. The argument list length is variable,
                but must be at least 2.

        Returns:
            datasketch.MinHash: A new union MinHash.
        '''
        if len(mhs) < 2:
            raise ValueError("Cannot union less than 2 MinHash")
        num_perm = len(mhs[0])
        seed = mhs[0].seed
        if any((seed != m.seed or num_perm != len(m)) for m in mhs):
            raise ValueError("The unioning MinHash must have the\
                    same seed and number of permutation functions")
        hashvalues = np.minimum.reduce([m.hashvalues for m in mhs])
        permutations = mhs[0].permutations
        return cls(num_perm=num_perm, seed=seed, hashvalues=hashvalues,
                permutations=permutations)

    @classmethod
    def bulk(cls, b, **minhash_kwargs):