Hi, How can I import some files for a node? I have a set of helper functions for my data science

node.py

and instead of having the helper functions within the same

node.py

, I want to store it in a separate

helper.py

in the same data science folder. I tried importing the functions from helper.py with

from .helper import funcHelp1, funcHelp2

but this makes the whole data science pipeline to be undiscoverable.

This looks like the right way to do this. Pipelines are not discovered if there is some kind of syntax or runtime error in the definition. Kedro auto-discovery makes this tricky to debug though because it fails behind the scenes. To see the error explicitly then, just import the pipeline manually in the

pipeline_registry.py

file. For example:

from my_pipeline import create_pipeline

pipe = create_pipeline()

Do

kedro run

and you should see a helpful error. Then delete those debug lines of code once you fix it.

Hey Afiq, it works unless there’s something wrong with the helper file, for example if imports for functions used by the helper functions are missing in

helper.py

?

A better long-term solution than the simply "quick fix/debug" above is to learn and use different linting and testing tools that will help you discover these errors quicker 🙂 A good place to start would be both

ruff

and

pytest