LSHash
LSHash copied to clipboard
kayzhu
A fast Python implementation of locality sensitive hashing.
I will continue to maintain LSHash in https://github.com/guofei9987/pyLSHash
I am trying to install this package. when I run pipe install it gives me this error: pip install lshash==0.0.4dev Collecting lshash==0.0.4dev Using cached lshash-0.0.4dev.tar.gz Complete output from command python...
Uses `PyScaffolding` to add a `setup.cfg` file and eliminate `__author__`, `__version__` and other constants from the `__init__.py` file so that `setup.py` doesn't have to `import lshash` before it has been...
lshash.py line 298 like that: @staticmethod def euclidean_dist(x, y): """ This is a hot function, hence some optimizations are made. """ diff = np.array(x)-y return np.sqrt(np.dot(diff, diff)) ---------------------------------------------------------------------- I change...
pip installed lshash and attempted to import LSHash however I get the above error. versions: python - 3.6.7 numpy - 1.15.4 I am operating within a conda environment.
Sorry to bother. This might be a stupid question. I really don't know how to set the hash size, is that depend on my data size? In the quick start...
i have installed redis successfully and on providing the argument storage_config=`{"redis": {"host": 'localhost', "port": 6379}}`,matrices_filename="/home/username/filename.npz" where filename.npz has not been created nor is hashtable stored on redis .. the program...
Many LSH implementations use Jaccard similarity to return matching result above a certain threshold say 80% match. Is possible to implement the same in this library.
when I set the LSHash's input_dim with 300 dimension. the result of lsh.query is [ ].
