Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Copying swizzles #609

Closed
mtsokol opened this issue Jun 24, 2024 · 4 comments · Fixed by #610
Closed

Copying swizzles #609

mtsokol opened this issue Jun 24, 2024 · 4 comments · Fixed by #610
Labels
enhancement New feature or request

Comments

@mtsokol
Copy link
Collaborator

mtsokol commented Jun 24, 2024

Hi @willow-ahrens,

I'm writing with one question regarding Tensor and SwizzleArray.

When I want to convert one format to another we have:

jl.swizzle(
    jl.Tensor(storage.levels_descr._obj, tensor._obj.body), *order
)

Which copies Tensor to the new format (and discards order of unwrapped tensor and sets order). This requires that order argument matches tensors's order.

I can also pass tensor's SwizzleArray that returns a Tensor with default F-major order:

jl.Tensor(storage.levels_descr._obj, tensor._obj)

I would like a new copy_swizzle(dest::SwizzleArray, src::SwizzleArray) function, which returns a SwizzleArray (or Tensor(dest::SwizzleArray, src::SwizzleArray) if it's more coherent).

This would allow me to pass CSC tensor and get a CSR array easily:

copy_swizzle(swizzle(Dense(SparseList(Element(0))), (2, 1)), csc_swizzle_array)

WDYT?

@mtsokol mtsokol added the enhancement New feature or request label Jun 24, 2024
@willow-ahrens
Copy link
Collaborator

willow-ahrens commented Jun 24, 2024

I think you can use copyto! here

@mtsokol
Copy link
Collaborator Author

mtsokol commented Jun 24, 2024

@willow-ahrens Thank you! With copyto! resetting order is avoided. But I noticed that copyto! from dense to sparse format treats zeros as non-fill-values. How can I avoid it?

using Finch

arr = [0 1 0; 2 3 0]
tns = Tensor(arr)
sw = swizzle(tns, 1, 2)

copyto!(swizzle(Tensor(Dense(SparseList(Element(0)))), 1, 2), sw)

SwizzleArray (1, 2)
└─ 2×3-Tensor
   └─ Dense [:,1:3]
      ├─ [:, 1]: SparseList (0) [1:2]
      │  ├─ [1]: 0
      │  └─ [2]: 2
      ├─ [:, 2]: SparseList (0) [1:2]
      │  ├─ [1]: 1
      │  └─ [2]: 3
      └─ [:, 3]: SparseList (0) [1:2]
         ├─ [1]: 0
         └─ [2]: 0

@mtsokol
Copy link
Collaborator Author

mtsokol commented Jun 24, 2024

I also managed to avoid copying fill-values with dropfills! but for csr format it fails with:

ERROR: Finch.FinchProtocolError("SparseListLevels cannot be updated multiple times")
using Finch

arr = [0 1 0; 2 3 0]
tns = Tensor(arr)
sw = swizzle(tns, 1, 2)

dropfills!(swizzle(Tensor(Dense(SparseList(Element(0)))), 2, 1), sw)

@willow-ahrens
Copy link
Collaborator

note: I have made a fix so that dropfills works like copyto. However, dropfills will always make a copy. You must still check for the cases where you can zero-copy during asarray.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants