Hi all, I am developing a modular namespace data science pipeline very similar to the ML example shown in this documentation: https://docs.kedro.org/en/0.18.2/tutorial/namespace_pipelines.html. In the latest documentation this is not covered. For 0.18.7 is this no longer a recommended design pattern? Also, when using namespace pipelines is it possible to have global parameters that apply to all pipelines? For example if I had several ML pipelines that I wanted to train with the same

train_start

and

train_end

time I could set a global parameter rather than duplicating these for each namespace? If not is there a way to replicate this functionality? Thanks a lot for your time, Andrew

1. No namespacing is still very much recommended, we just overhauled the tutorial based on feedback that it was too complicated. 2. Not currently, I’m personally very keen to improve the experience here

Thanks for the quick response @datajoely. I’ll proceed with a design similar to the namespace example. It’s very neat. Regarding 2 - that’s good to know. Hopefully that functionality will be available in the future to minimise duplication. Cheers