4 ms·
Definitely checking this out today! I use postgres for ~30 GB of machine learning data (object detection) and have a couple workflows which go through the Postg
by leighleighleigh 2y ago
Definitely checking this out today! I use postgres for ~30 GB of machine learning data (object detection) and have a couple workflows which go through the Postgres->Parquet->DuckDB processing route.
A couple questions, if you have time:
1. How do you guys handle multi-dimensional arrays? I've had issues with a few postgres-facing interfaces (libraries or middleware) where they believe everything is a 1D array!
2. I saw you are using pg_duckdb/duckdb under the hood. I've had issues calling plain-SQL functions defined on the postgres server, when duckdb is involved. Does BemiDB support them?
Thanks for sharing, and good luck with it!
- exAspArk 2y agoThank you, please give it a try! Great questions: 1. We currently don't support multi-dimensional arrays, but we plan to add support for such complex data structures. 2. Would you be able to share what type of user-defined functions are these, do they do modify the data or read it?
- leighleighleigh 2y ago1. good to hear! 2. The bulk of them are convenience wrappers which resolve UUIDs into other values, so most are read-only with only a single table lookup.