Hi all I would like to create a custom dataset which extends

pandas.GBQTableDataset

as I want the table schema to be generated automatically based on a polars DF. One challenge that I am facing is that GBQTableDataset expects a kwarg `dataset`, however in the catalog I want to use

kwarg

dataset to point kedro to the location of my custom dataset class. I have tried to rename the original kwarg

dataset

to

gbq_dataset

extend

pandas.GBQTableDataset

and then do

super().__init__(dataset=gbq_dataset)

However construction then fails with

GBQTableDataset.__init__() got an unexpected keyword argument 'gbq_dataset'

In this specific case, how can I point kedro to my custom dataset class while maintaining the original

dataset

parameter?

Hi Janick, Thank you for your question. What versions of Python and Kedro are you using?

kedro: 0.19.8 python: 3.11

define

type:

in the catalog pointing to my custom dataset

Does this solve your problem ?

Yeah I think so, just testing it