Is your feature request related to a problem? Please describe.
Vector usage is very skyrocketing now. so, market wants to handle with them using database. but we didnt. exactly, the vectorDB has many popularity from resource 'db-engine rank'.

related to a problem

but , Apache age doesn't have the function for vector engineering. so many other person who wants to apply the vector are just using the python library itself for supporting vector handling.

Describe the solution

Extension PGvector.

CRETATE EXTENSION vector;

SELECT embedding <-> '[3,1,2]' AS distance FROM items;
SELECT (embedding <#> '[3,1,2]') * -1 AS inner_product FROM items;
SELECT 1 - (embedding <=> '[3,1,2]') AS cosine_similarity FROM items;

and also can indexing function for searching of nearest neighbor. There are many very good functions except for the aforementioned functions.

Additional context
Except for above problem, we expand this feature to ML practitioner who wants data efficient management. the overview of additional context is compatible with Pytorch_geometric remote backend function. They builds the class for ease to integration others database. if previous problem well solve , then we might do this future work.

in below reference , it has the function for convenience.

[Scaling Up GNNs via Remote Backends]
https://pytorch-geometric.readthedocs.io/en/latest/advanced/remote.html