search

mljar / mljar-supervised

Python package for AutoML on Tabular Data with Feature Engineering, Hyper-Parameters Tuning, Explanations and Automatic Documentation
https://mljar.com
MIT License
3k stars 401 forks source link

figure legend shows every folds #137

Closed shahules786 closed 4 years ago

shahules786 commented 4 years ago

While plotting the learning curve for the neural network ( iterations > 1) , the figure shows legend as every fold instead of the specific fold which for which the results are recorded.

Code to generate the plot a = AutoML(total_time_limit=60,mode='Compete')

learning_curves

pplonski commented 4 years ago

Probably on other folds, there is only one iteration, that's why you don't see it on the plot. Could you try to increase the time limit, maybe to 10 minutes (10*60), time limit is in seconds.

BTW, @shahules786 how have you found this package?

shahules786 commented 4 years ago

I was doing a lot of research about opensource auto ml packages that are under active development to start contributing. Found this on some website. I think it will be awesome if you can start a slack channel to help contributors kickstart and interact with each other. @pplonski

shahules786 commented 4 years ago

Probably on other folds, there is only one iteration, that's why you don't see it on the plot. Could you try to increase the time limit, maybe to 10 minutes (10*60), time limit is in seconds.

Yes,you're right.

pplonski commented 4 years ago

@shahules786 good idea, I will set a slack! :)

pplonski commented 4 years ago

The slack link https://mljar-supervised.slack.com/join/shared_invite/zt-gkhfsvhw-H6LMKxxV5adeTmn9V7nbZw#/ I will add it to README.md

I'm closing this issue.