How to use scattered data in training process with DeepXDE?data

[rahaarbasi]

Dear Professor Lulu @lululxvi,

I am trying to use DeepXDE to solve a PDE with initial and boundary conditions and collocation points. I also have scattered data that I want to use in the training process. However, I am having trouble incorporating the scattered data into the training process.

I tried using the train_x and train_y arguments when defining my training data with the dde.data.PDE class, but it didn't work. Could you please provide some guidance on how to use scattered data in the training process with DeepXDE?

Thank you for your help!

[lululxvi]

See inverse problem examples.