EDMD doesn't seem to work for PDEs

[hannesiwr-droid]

Hello everyone,
I have spent the last few days trying to apply EDMD with polynomial basis functions to the tutorial example of Burgers' equation. The K-matrix is successfully created and the eigenvalues are displayed, but the "simulate"-method throws an error when entering the same initial data u0 as for the working DMD example. I have tried reformatting the initial data, but that did not work either. At the moment, it seems obvious that the implemented EDMD methods are not supposed to be applied to PDEs, but I would very much enjoy being able to.
Thank you!

[pswpswpsw]

thanks for reporting the error. can you post your code/notebook to reproduce the error? thanks. also the environment as well.

[hannesiwr-droid]

Thank you!
I've been using Jupyter on a Python 3.10 environment, using scikit-learn==1.0.2.
Here is my notebook: KoopmanExperiments.ipynb I first introduce the DMD model, with which there are no issues, and then later set up the (polynomial) EDMD model in cell 5. It succeeds in computing the K-matrix and its eigenvalues in cell 6, but the "simulate" function does not seem to work in cell 9. (Don't worry about cell 8 failing, it's for the DMD case). This failure appears due to the different data shape this EDMD model accepts. I have tried recasting the data into a form it will take, but I couldn't get it to work.