Hey team, I am working with a codease that uses kedro 0.18.3 and cannot change it. I need to run a pipeline through a python code iterating over some parameters. However, running the way I researched does not updtate the params. Is there something in my code I might be missing:

bootstrap_project(PROJECT_ROOT)


class TargetShare(TypedDict):
    min: float
    max: float


@supress_logs
def run_kvi_engine_with_inputs(
    kvi_target: TargetShare, foreground_target: TargetShare
) -> pd.DataFrame:
    with KedroSession.create(project_path=PROJECT_ROOT) as session:
        context: KedroContext = session.load_context()
        catalog: DataCatalog = context.catalog

        updated_params = context.params | {
            “kvi_engine.product_group_target_boundaries”: {
                “kvi”: kvi_target,
                “foreground”: foreground_target,
            }
        }

        catalog.add(“parameters”, MemoryDataSet(updated_params), replace=True)
        context._extra_params = updated_params
        session.run(pipeline_name=“kvi_engine”)
        rank_output = catalog.load(“kvi_engine.kvi_rank_output”)
        #rank_output is not changing
    return rank_output

Someone will reply to you shortly. In the meantime, this might help:

Hi @Daniel Mesquita, Can you try

add_feed_dict

- https://docs.kedro.org/en/0.18.9/kedro.io.DataCatalog.html#kedro.io.DataCatalog.add_feed_dict ?

Awesome. With this I can add the parameters in data catalog, however when I do session.run(pipe) in the sequence it does not receive the updated catalog. Is there a way to link it

Interesting, ideally the same kedro session should give you the updated catalog. Let me confirm. Thank you

is it version 0.18?

sorry it is 0.19.10 but let me give a try with 0.18.

Must be a problem in my end. Ended up writing a script to rewrite the yml 😆