search

c4dt / predictive-maintenance

An experiment to decentralize learning for predictive maintenance
0 stars 0 forks source link

Exp: using Linear Regression after Resnet #6

Closed lanterno closed 12 months ago

lanterno commented 1 year ago

TODO: add some explanation here. Avoid PRs with an empty PR message, even for draft PRs.

Closes #5

PascalinDe commented 1 year ago

I didn't look any further into the details at the moment, but I'd run a Python linter (e.g. Flake8 should be able to find unused variables such as abnormal_indices) on it to find other cases of code such as the one I pointed out

ineiti commented 1 year ago

I didn't look any further into the details at the moment, but I'd run a Python linter (e.g. Flake8 should be able to find unused variables such as abnormal_indices) on it to find other cases of code such as the one I pointed out

This is highly experimental code, so we should not spend too much time making it nice. It should be optimized for

PascalinDe commented 1 year ago

in my opinion the generated code is not well-written (cf. my comment) in the sense that the AI took some weird twists and turns (such as introducing unnecessary variables) which makes it hard to read and review

also, no docstrings, etc. in the code at all

while I personally don't mind if we use AI-generated code, I have some questions about how to review it, especially if it is more tedious than it would be with human-generated code

to elaborate: should I point out odditities? or not?