[FEATURE] DataFrames as a first class schema

[mynameisvinn]

🚨🚨 Feature Request

• Related to an existing Issue
• A new implementation (Improvement, Extension)

Should DataFrames be treated as a first-class schema (alongside Text, Audio, Images)?

If your feature will improve HUB

Even though Hub is designed for unstructured datasets, many data science projects still rely on DataFrames in some form or fashion. While there may be better solutions for large scale tabular data analysis, Hub should aim for feature parity.

Ultimately, the goal is to create a package familiar to data scientists, and supporting DataFrames is a big step towards that goal.

Description of the possible solution

Since a DataFrame is a collection of numpy arrays, a DataFrame schema could be assembled with existing Schemas (eg Tensors).

[mynameisvinn]

Closing this feature request due to inactivity and lack of interest. Will revive it if more users request it.

[mynameisvinn]

Closing this feature request due to inactivity and lack of interest. Will revive it if more users request it.