There is an unfortunate overloaded of the terms relational and relationships in relational databases.
Relationships is association between relations/tables, parent-child, node/edge etc, depending on model, extensions etc..
There are three basic models of databases:
model name. | basic data structure
-----------------------------------
relational | tables
hierarchical | trees
network | graph
A "relational" in RDBMS and Codd's rules is just a table data structure with some additional rules.
Part of those rules are a named table, with named and typed attributes (columns) with data in the form of rows of tuples.
PgVector is nearest neighbor search for tuple values, often from a single table/relation while PgRouting is graph traversal for relational data.
There is a bit more to that and in the relational model the data is independent of the schema, and no RDBMS is pure.
It is possibly helpful to realize that pgvector is about finding neighbors among tuples, in that relation/table, and that it is very different from graph traversal.
Relationships is association between relations/tables, parent-child, node/edge etc, depending on model, extensions etc..
There are three basic models of databases:
A "relational" in RDBMS and Codd's rules is just a table data structure with some additional rules.Part of those rules are a named table, with named and typed attributes (columns) with data in the form of rows of tuples.
PgVector is nearest neighbor search for tuple values, often from a single table/relation while PgRouting is graph traversal for relational data.
There is a bit more to that and in the relational model the data is independent of the schema, and no RDBMS is pure.
It is possibly helpful to realize that pgvector is about finding neighbors among tuples, in that relation/table, and that it is very different from graph traversal.