GOMODULES
/
hyper
mirror of https://github.com/vitali-fedulov/hyper.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
Hashing float vectors in N-dimensions
41 Commits
1 Branch
7 Tags
73 KiB
Go 100%
 
Go to file
Vitali Fedulov bd63fbbcd4 fnv.New64a test fix and README update 2024-02-07 20:55:42 +01:00
LICENSE fnv.New64a test fix and README update 2024-02-07 20:55:42 +01:00
README.md fnv.New64a test fix and README update 2024-02-07 20:55:42 +01:00
cubes.go fnv.New64a test fix and README update 2024-02-07 20:55:42 +01:00
cubes_test.go switch to methods 2022-01-19 09:08:56 +01:00
go.mod first commit 2021-10-25 15:50:52 +02:00
hashes.go switch to methods 2022-01-19 09:08:56 +01:00
hashes_test.go fnv.New64a test fix and README update 2024-02-07 20:55:42 +01:00

README.md

Hashing N-dimensional float vectors

Search nearest neighbour vectors in n-dimensional space with hashes. There are no dependencies in this package.

Each vestor is discretized into a set of hashes, as described here (also as PDF).

How to use

  1. Provided a float vector []float64, use CubeSet and CentralCube functions to generate hypercube coordinates []int. The difference between the two functions is that one corresponds to hash-table record and the other to a query or vice versa, depending on performance/memory preference.
  2. HashSet and DecimalHash/FNV1aHash are used to get corresponding hash set and central hash from the hypercube coordinates above. There are 2 alternative hash functions: DecimalHash and FNV1aHash. DecimalHash does not have collisions, but is not suitable for cases with large number of buckets or dimensions. FNV1aHash is applicable for all cases.

Example for similar image search/clustering.

Go doc for full code documentation.