Hi Everyone. I only recently started using Dagster so I have a somewhat ideological question. Is DataFrame a central part of a Dagster architecture? Is Dagster still a good fit for my use cases in a case if I plan to reduce dependency on pandas as much as possible?

Hi Nikita, Dagster itself has no concept of “DataFrame” so no, it’s not a part of the Dagster architecture. However, Dagster is intended to be agnostic with respect to your data structures. Certainly we expect many of ours users to be processing tabular data, including in data frames. So we provide several integrations and UI niceties specifically geared towards pandas and tabular data, but it’s also not difficult to roll your own solution. I suggest having a look at this guide on using our integration dagster-pandera

Thanks a lot, I'll be looking into it