An interface with Pandas

[simonaubertbd]

Problem Statement

Hello,
As of today, you can convert easily a string. But what I want to do, as a contribution to Amphi ETL is to interface it with Pandas.

Proposed Solution

Several use case, either to or form toon.

pandas_to_toon would be like an an aggregation, given a pandas string (the table represents a dataframe). Example :

Customer	Customer Country	Customer question
1	FR	How do you do?
2	UK	How old are you?
2	UK	Are you hungry?

could give, by row (here, I choose only Customer and question)

Customer	Customer Country	Customer question	Generated Toon
1	FR	How do you do?	Customer: 1 "Customer Question": How do you do?
2	UK	How old are you?	Customer: 2 "Customer Question": How old are you?
2	UK	Are you hungry?	Customer: 2 "Customer Question": Are you hungry?

But I can also have

Customer	Customer Question	Generated Toon
1	FR	Customer: 1 "Customer Question": How do you do?
2	UK	Customer: 2 "Customer Question": mydataframe[3]{Customer,"Customer Question"}: 2,How old are you? 2,"Are you hungry? "

or even

Generated Toon
[3]{Customer,"Customer Question"}: 1,How do you do? 2,How old are you? 2,"Are you hungry? "

Alternatives Considered

None

SPEC Compliance

No response

Additional Context

I would like to use it on Amphi-ETL in addition with an LLM