Should julia libs adhere to a common Array & dataframe API?

General Usage
,
1

Hi,
I recently discovered this: Announcing the Consortium for Python Data API Standards. This is of course for Python but I wonder if a similar effort in julia would be opportune.

I realise that in julia due to multiple dispatch a lot of libs are already adhering to a common API, especially for arrays but for dataframes the issue might be less settled. Maybe inspiration for a good common dataframe api could be gotten by looking at what these guys come up with?

For one thing it might help with interop with external libs/systems like Spark, array databases, etc. in other languages that may adhere to this standard in the future.

2

There already exists something similar to what you mentioned, called Tables.jl. There is also an effort for arrays called ArrayInterface.jl.

2 Likes
3

Good stuff