Should Pandas be added as requirement for nebula-python?

[magnudae]

Should pandas be added to the dependency list?

I am playing around with the ResultSet in my project at the moment and I stumbled upon that if I use as_data_frame I get the error I need Pandas python.

I thought this would automatically come with nebula-python since its a dependency, but then I see its not in the dependency list of the project.

Should pandas be added to the dependency list? I see it comes down in the lock file here, but that only declares the lockfile for this project (not projects depending on this)

[wey-gu]

Thanks @magnudae , it's on-purposely not included as dep to keep nebula-python slim as the sugar of as_data_frame was considered not a mandatory user path, but we could consider adding optional profile like pip install nebula3-python[dataframe] to improve dx? what do you think?

[magnudae]

Thanks @magnudae , it's on-purposely not included as dep to keep nebula-python slim as the sugar of as_data_frame was considered not a mandatory user path, but we could consider adding optional profile like pip install nebula3-python[dataframe] to improve dx? what do you think?

Yeah after looking at it for a little bit more, I agree with you that it should be a core dependancy, as you say, to keep it slim.

Adding the [dataframe] would be helpful, but I think it shouldnt be prioritised over other things

thank you @wey-gu for the answer, ill close my question here since I think it is rightfully not included in the dependency list