Open okennedy opened 2 years ago
Spark has some pretty good ETL capabilities built in. It might be fairly easy to add a cell or similar construct that lets us use them to build ETL pipelines.
https://spark.apache.org/docs/latest/ml-features.html
Side note: Would it be worth reframing Vizier Dataset artifact objects as spark transformers?
This should actually be a purely UI-based challenge, now that we support Transformers as Vizier datasets.
Spark has some pretty good ETL capabilities built in. It might be fairly easy to add a cell or similar construct that lets us use them to build ETL pipelines.
https://spark.apache.org/docs/latest/ml-features.html
Side note: Would it be worth reframing Vizier Dataset artifact objects as spark transformers?