-
E.g., Suitesparse's `GxB_LAND_UINT32`
-
As per this comment: https://github.com/metagraph-dev/mlir-graphblas/pull/121#discussion_r694568857
We can use `clang-format` to format our C++ in the style used by MLIR. We should do this once to…
-
How do the operators behave with Nan and Inf? what is 'max(x,NaN)'?
See 'fmax' in ANSI C11, which is the "omit nan" behavior, so that 'max(x,NaN)=x' for all 'x'. That is faster since it allows MAX …
-
Currently, we have many functions like `csr_densify5x5` and `csc_densify8x8` in many of our tests.
Since many of our test files use their own JIT engine fixtures whose scope is just the test file,…
-
I've not tested yet in C but in both Rust and Java when running `GrB_Matrix_apply` with unaryOp `one` I get `0` instead of `1`
Steps to reproduce:
```rust
fn test_matrix_apply_transpose_on_self_f…
-
Keep track of missing SparseArrays function(alitie)s:
- [ ] sprand
- [ ] sprandn
- [ ] Some alternative to sparse
- [ ] spzeros(?)
- [ ] spdiagm
- [ ] blockdiag
- [ ] rowvals
- [ ] nzrange
…
-
Hi, This s really a question rather than a bug, but I have a hard time understanding the documentation of how to compile suitesparse with MKL
The ReadMe it says you can do
```
make BLAS=-lmyblasl…
-
I define GxB_INDEX_MAX as 2^60, which defines the largest dimension of any matrix or vector I can allow. UINT64_MAX is too big. I suggest that each library defines its own GrB_INDEX_MAX.
-
Here's the inevitable question ;)
What are the differences (philosophical, implementation etc) between this and geometric flux?
Are you covering a smaller scope? I think graphs are a subset of g…
-
Right now the FeaturedGraph type has arbitrary underlying storage (e.g. adjacency_matrix, adjacency list, ...).
This is not ideal, since gnn layers have to perform additional computations (e.g. produ…