Open aadya940 opened 3 months ago
maybe add some context as to why?
@TonyBagnall @aadya940 Would you be able to provide more context on this? @aadya940 Is it open for me to solve?
@av6-github Yes, go ahead :) Basically, you have to remove the following layer from the clusterer's since this is never actually needed.
https://github.com/aeon-toolkit/aeon/blob/main/aeon/clustering/deep_learning/_ae_fcn.py#L232
@aadya940 thank you!
@aadya940 since there is no output layer variable, what do I put in the outputs parameter in the model?
@av6-github You have to pass the decoder_output
layer.
@aadya940 i am having some issues with pre commit and certain checks, so i am currently unable to make a proper pull request. :( Apologies!
@av6-github Hello, your PR seemed good to me. If you're having issues with Pre-Commit here are the steps:
Install:
python -m pip install pre-commit
Run:
pre-commit run --files path_to_file/file.py
Do let us know if you need any help.
@av6-github do you still need a hand with this? It is fine to keep a pull request open even if it is failing the CI currently.
@av6-github do you still need a hand with this? It is fine to keep a pull request open even if it is failing the CI currently. Oh, thank you so much for the consideration, it means a lot! I am really new to this so am a little confused here! It'd be great if you could help me out here!
@av6-github Hello, your PR seemed good to me. If you're having issues with Pre-Commit here are the steps:
Install:
python -m pip install pre-commit
Run:
pre-commit run --files path_to_file/file.py
Do let us know if you need any help.
I tried this, but still am having issues! Should I fork the repo again and do the changes on that and open a new pr?
Not sure what the issue is without logs really. You can check the errors by creating a PR and then clicking on details
on the failing test.
hi @av6-github we were just reviewing open issues, are you still interested in resolving this? If so, @aadya940 could you advise? thanks.
Describe the issue
The
tf.keras.layers.Reshape
was introduced inclustering/deep_learnin
to make the input shape of the encoder equal to the output shape of the decoder which is already being tested intest_all_networks.py
. So, basically it is unnecessary at the moment.Suggest a potential alternative/fix
Remove it.