Is it possible to share the JSON file that you have? or mask any sensitive information. Also which kedro, and kedro-viz version are you using?

yeah of course, and im using kedro viz verson 5.1.1 and kedro version 0.18.2

message has been deleted

I cannot reproduce the error, everything works fine on my side.

Can you copy the full error trace here? I need to check if the error is coming from kedro or kedro-viz, since the screenshot you post suggest that something went wrong when it is iterating the path.

message has been deleted

Does this give you the same error?

Before we create a Kedro session, we need to change directory to the location where the the pyproject.toml file and the conf files are located right? Is there anything else it needs to create the Kedro session?

You can tweak the pyproject.toml to choose a diff directory iirc

I am using dbx to deploy my code to databricks and with each deployment the code is deployed to a uniquely regenerated uuid folder. So it is hard for me to tell Kedro where the pyproject.toml file is located as with each deployment the project root changes.

I am not sure how dbx works, but how would you execute code from there for non-kedro project? I imagine you must start from the same directory, why do you need to specify the absolute path?

Hello guys, How can i get the dataset name, specidfied in the catalog, from a dataset class for example GBQQueryDataSet?

Hi Kedro team, do you have some guidelines as per when is it better to split one node into multiple nodes, and conversely when is it better to merge multiple nodes into one? For example, i am working on a feature engineering pipeline that operates on pandas dataframes, and it's not exactly clear for me when i should stack a whole bunch of transformations/augmentations in one single node, or instead split that code into multiple nodes. Would you have some feedback on this? Thanks!

Custom DataSet

We typically use the heuristic "what's easy to test". That leads us to often break nodes down into small single purpose parts rather than big monoliths.

I should say we explicitly mean unit and integration test here

ah ok thanks, so I guess you have unit test for every nodes in your pipelines? I don't have too many unit tests at the moment, and when I have I test more at the output of a pipeline level

Yeah you can confidently test the function of a unit of logic with simple data

Hey, is it possible to use Kedro's way of handling configuration (paths, envs, etc.) for training a docker image with Sagemaker (from within a pipeline)?

hi I tried to develop kedro plug-in locally. When I obtained the pipeline, I found that the node tags were not obtained. But I did when I defined it

Hey guys, how do I prevent a specific node from running until an event happens?

We do typically encourage this because it introduces combinatorial complexity. But env vars are the easiest way imo

thanks, sorry for the lazy question, too much on my head, and too early in the morning...

Hey team, what best approach do you have to generate a data catalog from catalog.yml?

Generate catalog from catalog.yml

These are dedicated outputs to a pipelines, so they are not namespaced in their origin pipeline (or the catalogue). But now these origin pipelines are part of a modular pipeline . Now that I wrapped these pipelines inside a Modular pipeline they've become namespaced and roam around my kedro viz, if I don't unfold the pipelines. What is the best way to handle this?

Hello, In the spaceflight tutorial, the following code is used to create a pipeline:

python
from kedro.pipeline import Pipeline, node, pipeline
def create_pipeline(**kwargs) -> Pipeline:
    return pipeline(
        [
            node(),
            node(),
        ]
    )

I have a "Module is not callable" lint error due to the use of

pipeline()

(without a capital letter). The message disappears when I replace

pipeline()

by

Pipeline()

. What is the difference between these 2? and why are we importing both?

Does it gives you an error when you run it? pipeline is a function that helps you to create the Pipeline object.