| pyvgx Similarity Members | Description |
|---|---|
Maximum hamming distance allowed between vector fingerprints to be declared "similar" |
|
Minimum similarity score allowed between vectors to be declared "similar" |
|
Relative importance of Cosine() in Similarity() score computation |
|
Relative importance of Jaccard() in Similarity() score computation (Feature Vector only) |
|
Minimum Cosine() allowed between vectors to be declared "similar" |
|
Minimum Jaccard() allowed between vectors to be declared "similar" (Feature Vector only) |
|
Minimum overlapping dimensions allowed between vectors to be declared "similar" (Feature Vector only) |
|
Maximum number of elements in vectors |
|
Set of pre-defined ANN seeds |
|
Future. |
|
Future. |
1. pyvgx.Similarity.hamming_threshold
Maximum hamming distance allowed between vector fingerprints to be declared "similar"
2. pyvgx.Similarity.sim_threshold
Minimum similarity score allowed between vectors to be declared "similar"
3. pyvgx.Similarity.cosine_exp
Relative importance of Cosine() in Similarity() score computation
4. pyvgx.Similarity.jaccard_exp
Relative importance of Jaccard() in Similarity() score computation in Feature Vector mode.
| Must be 0.0 in Euclidean Vector mode |
5. pyvgx.Similarity.min_cosine
Minimum Cosine() allowed between vectors to be declared "similar"
6. pyvgx.Similarity.min_jaccard
Minimum Jaccard() allowed between vectors to be declared "similar" in Feature Vector mode.
| Ignored in Euclidean Vector mode |
7. pyvgx.Similarity.min_isect
Minimum overlapping dimensions allowed between vectors to be declared "similar" in Feature Vector mode.
| Must be 1 in Euclidean Vector mode |
8. pyvgx.Similarity.max_vector_size
Maximum number of elements in vectors
9. pyvgx.Similarity.seeds
Internal set of pre-defined seeds for use with ANN search functionality. This is a special case internal implementation which is subject to change, and is thus intentionally undocumented.
10. pyvgx.Similarity.nsegm
Future.
