CODEWITHSUNDEEP
tensorflowpython-3.6linear-algebrachainercupy
Nacho
Nacho

Reputation: 832

Solve rectangular sparse linear equation systems with cupy

I'm trying to solve a rectangular system for sparse features using cupy. I know the built-in function sparse.linalg.lsqr(A, b) do it for square matrix A. However I like to solve a rectangular sparse system. This is the way we can solve a squared system: 

Import cupy as cp

A = cp.sparse.rand(200, 100, density=0.1)
b = cp.random.random(100)

x = cp.sparse.linalg.lsqr (A, b)
print(x)

It gives an error of dimension mismatch for rectangular systems and I can't find a built-in sparse method equivalent to e.g. cupy.tensorsolve().

By the way, is there a way to do it with Tensorflow? Thank you for help. I'm using google a Colaboratory notebook.

Upvotes: 2

Views: 575

Answers (2)

Yuki Hashimoto
Yuki Hashimoto

Reputation: 1073

It might be too late for questioner, but for posterity I answer this question.

This can be implemented by wrapping lsqr in MAGMA as shown in this example.

Upvotes: 2

corochann
corochann

Reputation: 1624

You can refer https://docs-cupy.chainer.org/en/stable/reference/linalg.html for the current cupy's supporting functions for linear algebra. I can find cupy.linalg.tensorsolve, but I could not find cupy.linalg.lsqr so far.

Also I think it is nice to tag this question as "cupy" or "numpy".

Upvotes: -1

Related Questions