TL;DR If you answer “yes” to any of these questions, this video is for you.
- Are you experiencing memory issues working with production level data in a development environment?
- Do you have to maintain two version of your SQL in development and production environments?
- Do you just want to write Python (and not SQL)?
Recently, PyData released video of a talk Gil Forsyth, a Software Engineer here at Voltron Data, gave at PyData NYC conference. The talk focuses on the value Ibis has for organizations by focusing on:
- “the translation problem,” which is the lack of standardization across SQL variants and;
- “the parameterization problem,” which is that queries can be structured in various ways depending on developer assumptions and variable inputs.
At around the 24-minute mark in the video, Gil shows how seamless it is to run the same Ibis code against three different database management systems — and he does this in about 30-seconds.
“...there’s nothing about this that says run on DuckDB. It just says run this on something that has columns that have these names and these data types… so if I have a different connection, say, to like a Postgres instance, I can execute the same thing… and if I wanted to say, have a third connection to a local SQLite database, then I can do the same thing so long as the underlying database has these column names and those types, you can execute this.”
It’s only a 30-minute talk and there are 15 minutes of Q&A in which Gil lays out some exciting developments to come (*hint: SQL —> Ibis Expression —> Python).
Cloud + Forsyth- Ibis- Expressive analytics in Python at any scale | PyData NYC 2022
(Embed Video)
If you are interested in using Ibis, Arrow, or any of the other open source systems that we support and would like some assistance, Voltron Data can help. Please contact us directly or check out our Enterprise Support options.
