Aaron Meyer

Implementation of tensor PLS

2

#### Is your feature request related to a problem? Please describe. This is just a note that @cyrillustan and @JacksonLChin will be working on an implementation of tensor PLS over...

A new random interface

2

Several backends have now implemented the [NumPy random generator interface](https://numpy.org/doc/stable/reference/random/generator.html). With this, we can rethink how random state is passed throughout the random module. It may also be helpful to...

Callback interface for CP decomposition

16

I'd like to start a discussion on a callback function in each method as this will be helpful to several efforts (e.g., #368, #372). Here, I just implemented the simplest...

Testing for complex tensors

1

See #298. Several of the tests for methods should be extended to the complex cases. Perhaps a first step to accomplishing this is to extend the random tensor methods to...

I think we may want to use the opportunity to add some documentation, perhaps in the user_guide, either in the [backend](https://github.com/tensorly/tensorly/blob/main/doc/user_guide/backend.rst) section or in a new page? We can there...

The issue in the tests comes from the fact that we changed the SVD interface in #429, the best would be to use `tensorly.tenalg.svd_interface`. @aarmey perhaps we should deprecate tl.partial_svd...

Adding linear algebra and other array operations

13

It looks like this is still missing many matrix operations like QR, SVD, einsum, etc. Is there a clear path to using these with or without MLX? [This has been...

Consistent convergence checks and options

7

Related to #470, I have been wondering if we should make the convergence handling consistent and reusable across methods. Currently, methods handle convergence checks in slightly different ways—some have an...