Pipeline Caching?

Technical Help
#1

Hi there,

When I reference the pipeline creation in an external file, like:

from codeload.quantrocket.buy_gap_zipline.pipeline import make_pipeline

any changes to the pipeline code don't seem to be reflected when I re-run the backtest algorithm.

When I include the pipeline creation inline within the strategy file, any changes are picked up, and the algorithm executes as expected.

Is there any caching that occurs when the pipeline is created in a file outside the strategy and referenced?

David

#2

The two places you can check for caching are in the __pycache__ and .ipynb_checkpoints folders in the same directory as the file. __pycache__ is managed by Python and .ipynb_checkpoints is managed by JupyterLab. In theory you shouldn't have to think about those folders but neverthless it's something you can check.