SageMaker Data Wrangler UI Features

Question

The SageMaker Data Wrangler UI in SageMaker Studio doesn't seem to support all the features that the API does.  When will the UI support:
* Loading all s3 objects under a prefix?  https://aws-data-wrangler.readthedocs.io/en/stable/stubs/awswrangler.s3.read_csv.html#awswrangler.s3.read_csv
* Loading JSON objects in addition to CSV and Parquet files?  https://aws-data-wrangler.readthedocs.io/en/stable/stubs/awswrangler.s3.read_json.html#awswrangler.s3.read_json

Accepted Answer

As mentioned by Tulio Alberto in comments, Amazon SageMaker Data Wrangler (the graphical data preparation feature inside Amazon SageMaker) is separate from AWS Data Wrangler (an open-source data prep utility published by AWS Labs): The two tools are based on different technologies and don't necessarily aim for full feature parity - they just happen to share similar names.

To my knowledge there's no committed timeline we can share at the moment for when these particular features will make it to SageMaker Data Wrangler, but I think as feature requests they make sense and the reasoning for both is pretty clear: I'm aware that both have been discussed to some extent internally already, and I'd personally like to see them launch too!

Thanks for sharing the feedback, and apologies for the naming confusion!

Answer

Hi Tulio, thanks for the clarification.  But doesn't SageMaker Data Wrangler generate code that complies with/uses AWS Data Wrangler?  Isn't there some (if tenuous) connection between the two?

Answer

Hi,

SageMaker Data Wrangler in Studio  just launched the JSON/ORC support and we support import files under a prefix already. Please see the following links

- https://docs.aws.amazon.com/sagemaker/latest/dg/data-wrangler-import.html#data-wrangler-import-s3
- https://aws.amazon.com/blogs/machine-learning/prepare-and-analyze-json-and-orc-data-with-amazon-sagemaker-data-wrangler/

SageMaker Data Wrangler UI Features

Contenido relevante