datajoely / modular-spaceflights Goto Github PK

View Code? Open in Web Editor NEW

10.0 10.0 4.0 18.75 MB

This is a prototype version of the Kedro spaceflights tutorial that utilises the modular pipelines pattern.

Python 1.12% Jupyter Notebook 98.87% Makefile 0.02%

modular-spaceflights's People

Contributors

Stargazers

Watchers

Forkers

studioswong martinlarsalbert febikammck nanaoka

modular-spaceflights's Issues

Is the parameter `derived_params` supposed to be a List type?

The input is:

  derived:
    - column_a: number_of_reviews
      column_b: total_fleet_count
      numpy_method: divide
      conjunction: over # {column_a}_{over}_{column_b}
    - column_a: number_of_reviews
      column_b: total_fleet_count
      numpy_method: divide
      conjunction: over
    - column_a: reviews_per_month
      column_b: company_rating
      numpy_method: multiply
      conjunction: by

So the type should be List[Dict[str, str]] instead of just Dict[str, str], right?

Pipeline describe/run doesn't work as expected

I'm currently exploring all the pipelines inside this particular tutorial.

kedro pipeline describe -v feature_engineering
kedro run --pipeline feature_engineering

These particular commands share an error stating that this particular pipeline does not exist.
Error: `feature_engineering` pipeline not found. Existing pipelines: [Data ingestion, Feature engineering, Modelling stage, Pre-modelling, Reporting stage, __default__]

I'm guessing that it is just because the pipeline has been named as "Feature Engineering" in the pipeline registry.

Could you please confirm?

Recommend Projects