Hi Kedro Team I'm using Kedro

0.16.6

and using the

load_context

to get

params

,

"credentials*", "credentials*/**"

. We're upgrading

Kedro

to

0.18.8

and it seems

load_context

is no more accessible. How can I replicate the same functionality in the Kedro 0.18.8?

Implement a hook

after_context_created

I ment ahead and looked at the various migrations guides at https://github.com/kedro-org/kedro/blob/main/RELEASE.md, and I don't see much practical advice. I'll be opening an issue, because I think having better guidelines there might benefit lots of users

Thanks very much for the inputs @Juan Luis, @marrrcin My use case is to retrieve values stored in

parameters

and

credentials

in a script, so

after_context_created

may not be what I'm looking for. I'll try to create a

KedroSession

and access

load_context

from there to get the

parameters

and

credentials

.

I'm not sure creating a new

KedroSession

will give you what you need, but I could be wrong. how are you using the script? is it something you call as part of a node?

The script is basically used to fetch parameters and credentials for initialising an

MLFlowClient

. This is one example, its being used for many other purposes also. An example below:

project_root = _def_get_parent_dir()

# Load kedro context in order to reuse parameters
context = load_context(project_root)

tracking_uri = context.params.get("mlflow_tracking_uri")
registry_uri = context.params.get("mlflow_registry_uri")

credentials = context.config_loader.get("credentials*", "credentials*/**")
mlflow_creds = credentials.get("mlflow")

client = MlflowClient(tracking_uri=tracking_uri, registry_uri=registry_uri)

I see - in this case, it might indeed make sense to create your own

KedroSession

Why aren’t you using

kedro-mlflow

? It’s a great, production tested plugin @Afaque Ahmad

Even if you don't want to use

kedro-mlflow

(which does something very similar to what you are trying to achieve), it is worth checking the code to see how this is handled: The idea is to store what you need through a hook https://github.com/Galileo-Galilei/kedro-mlflow/blob/845ad919c9dbd020e948e8adc2e0f9064de1ef68/kedro_mlflow/framework/hooks/mlflow_hook.py#L38 and then retrieve through

session.context

attribute. Check this line to see how the context is modified by the hook and theses discussions (1, 2, 3) to understand why and how things were modified along kedro versions updates. If you have the

kedro-mlflow

plugin installed, you can access the client through

context.mlflow.server._mlflow_client

.

This is super helpful, I'll refer to this while doing the migration and get back again if I have questions. Thanks again @Yolan Honoré-Rougé @Juan Luis @marrrcin