topic Re: Pyflake errors when using %run in Data Engineering

Pyflake errors when using %run

Eeg — Sun, 01 Oct 2023 03:14:27 GMT

I am using %run command to import shared resources for each of my processes. Because it was the most easy way to import my common libraries. However, in that way, pyflake can't resolve the dependencies quite well. And I end up working in code with many red lines populated by pyflake. I have attached just one pyflake dependency error.

Is there a way for pyflake to load code which been loaded using %run magic command? Or in general how are you guys handling common utility file aside from the main processing file inside databricks?

Re: Pyflake errors when using %run

Eeg — Sun, 01 Oct 2023 03:15:41 GMT

The visualization of the problem I am referencing:

Re: Pyflake errors when using %run

btafur — Tue, 03 Oct 2023 07:58:45 GMT

You could use something like flake8 and customize the rules in the .flake8 file or ignore specific lines with #noqa.

https://flake8.pycqa.org/en/latest/user/configuration.html

Re: Pyflake errors when using %run

Eeg — Thu, 05 Oct 2023 02:53:39 GMT

Thank you for your reply, it's definetaly a way not to see the red lines.

However I am not sure it's a feasible way.

It would be best if there's a way a code formatter could detect classes and imports from file which loaded to databricks by %run command.

Re: Pyflake errors when using %run

Eeg — Thu, 05 Oct 2023 07:05:37 GMT

Sure, is this the right way to raise a request?

https://docs.databricks.com/en/resources/ideas.html#submit-feedback-from-the-workspace