Remove code being deprecated in v0.39

[mudit2812]

[sc-73935]

• Remove deprecated code from demos.
• Update various from pennylane import numpy as np to import numpy as np.

[github-actions]

👋 Hey, looks like you've updated some demos!

🐘 Don't forget to update the dateOfLastModification in the associated metadata files so your changes are reflected in Glass Onion (search and recommendations).

Please hide this comment once the field(s) are updated. Thanks!

[mudit2812]

The build failure here seems to come from sphinx_gallery, and I don't know how to debug it, since there is no traceback. I've verified that the versions of pennylane and lightning that are installed account for all deprecations. Catalyst could be causing the issue, but qml has no references to qml.Device, which is the deprecated class that is causing the failure.

[josh146]

Hey @mudit2812, how come the demos were updated to use pnp instead of np?

While we have been using pnp internally in the source code, we have deliberately kept demos/tutorials using from pennylane import numpy as np to stay consistent with all of our public documentation, papers, slides, codebook, etc.

[mudit2812]

Hey @mudit2812, how come the demos were updated to use pnp instead of np?

While we have been using pnp internally in the source code, we have deliberately kept demos/tutorials using from pennylane import numpy as np to stay consistent with all of our public documentation, papers, slides, codebook, etc.

Thanks @josh146 , I'll revert the change. Though for the tutorial_tn_circuits demo, I can't get the demo to work unless we use vanilla numpy for the template_weights given to qml.MPS, and we need pennylane.numpy for optimization. What is the recommendation for demos where both vanilla and pennylane numpy are needed?

[josh146]

What is the recommendation for demos where both vanilla and pennylane numpy are needed?

I don't think we've been in a position where we've had to do this in a demo before 🤔

Though for the tutorial_tn_circuits demo, I can't get the demo to work unless we use vanilla numpy for the template_weights given to qml.MPS, and we need pennylane.numpy for optimization.

Interesting! do we know why MPS doesn't work with PL numpy (sounds like a bug?)

In the meantime, I think it's okay to import vanilla numpy here (since original numpy is only being used to patch a bug/issue, we could import it as import numpy as onp)

[github-actions]

Thank you for opening this pull request.

You can find the built site at this link.

Deployment Info:

• Pull Request ID: 1220
• Deployment SHA: d7a8ae7b48ee47ddfface8ad706ec30917a07b62
  (The Deployment SHA refers to the latest commit hash the docs were built from)

Note: It may take several minutes for updates to this pull request to be reflected on the deployed site.