pytorch_cluster

pytorch_cluster copied to clipboard

Variable K in KNN

[{"_id":"66d0f0db5a1d13a447096757","body":"Yeah, you are right on this one. Currently, we assume fixed `k` across batches to keep required buffers during computation constant. Changing this would probably require a separate function. An alternative could be to utilize `max(k)` across all batches, and then filter invalid nodes in a post-processing step, but this may be inefficient as well if your `k` differs highly across examples.","issue_id":1710340833875,"origin_id":1501080098,"user_origin_id":6945922,"create_time":1681030763,"update_time":1681030763,"id":1724969179623,"updated_at":"2024-08-29T22:06:19.623000Z","created_at":"2024-08-29T22:06:19.623000Z"},{"_id":"66d0f0db5a1d13a447096758","body":"This issue had no activity for **6 months**. It will be closed in **2 weeks** unless there is some new activity. Is this issue already resolved?","issue_id":1710340833875,"origin_id":1751535256,"user_origin_id":41898282,"create_time":1696639944,"update_time":1696639944,"id":1724969179628,"updated_at":"2024-08-29T22:06:19.627000Z","created_at":"2024-08-29T22:06:19.627000Z"}]

Thank you very much for developing such a great library. Currently, I have a problem that is bothering me. In KNN, if I want to choose the number of neighbors...

mrzzmrzz

[MPS] PyTorch Nightly with MPS on Mac M1 does not work

[{"_id":"66d0f10c5a1d13a447096761","body":"Hm, ok I might've just realised that one would have to specifically implement code for the MPS interfaces which are currently not implemented and might be not too easy to implement, so I guess I am going back to CPU based tests...\r\n\r\nDo I understand the issue correctly? If so the issue can be closed","issue_id":1710340833882,"origin_id":1491852666,"user_origin_id":34018356,"create_time":1680265624,"update_time":1680265624,"id":1724969228289,"updated_at":"2024-08-29T22:07:08.288000Z","created_at":"2024-08-29T22:07:08.288000Z"},{"_id":"66d0f10c5a1d13a447096762","body":"Yes, I am sorry. We don\u2018t have any MPS-specific implementation right now. You might be able to convert to CPU before calling the cluster routines, and back to MPS afterwards.","issue_id":1710340833882,"origin_id":1493245629,"user_origin_id":6945922,"create_time":1680417506,"update_time":1680417506,"id":1724969228371,"updated_at":"2024-08-29T22:07:08.371000Z","created_at":"2024-08-29T22:07:08.371000Z"},{"_id":"66d0f10c5a1d13a447096763","body":"This issue had no activity for **6 months**. It will be closed in **2 weeks** unless there is some new activity. Is this issue already resolved?","issue_id":1710340833882,"origin_id":1741609396,"user_origin_id":41898282,"create_time":1696035099,"update_time":1696035099,"id":1724969228383,"updated_at":"2024-08-29T22:07:08.382000Z","created_at":"2024-08-29T22:07:08.382000Z"}]

Hi, I am currently trying to get pytorch geometric to run on my Mac with M1 Pro, the recent update to 13.3 fixed quite a few things. Look here for...

nicolas-schreiber

General Union types are not currently supported in v1.6.1

[{"_id":"66d0f0f4b3dc7ee8eb11aa31","body":"What PyTorch version are you on?","issue_id":1710340833891,"origin_id":1737506614,"user_origin_id":6945922,"create_time":1695824575,"update_time":1695824575,"id":1724969204888,"updated_at":"2024-08-29T22:06:44.887000Z","created_at":"2024-08-29T22:06:44.887000Z"},{"_id":"66d0f0f4b3dc7ee8eb11aa32","body":"This issue had no activity for **6 months**. It will be closed in **2 weeks** unless there is some new activity. Is this issue already resolved?","issue_id":1710340833891,"origin_id":2019195256,"user_origin_id":41898282,"create_time":1711414411,"update_time":1711414411,"id":1724969204892,"updated_at":"2024-08-29T22:06:44.891000Z","created_at":"2024-08-29T22:06:44.891000Z"}]

RuntimeError: General Union types are not currently supported. Only Union[T, NoneType] (i.e. Optional[T]) is supported.: File "/home/zj/miniconda3/envs/yukewang/lib/python3.8/site-packages/torch_cluster/rw.py", line 18 num_nodes: Optional[int] = None, return_edge_indices: bool = False, ) -> Union[Tensor,...

Albert-Zhao-2020

Infinite loop in torch_cluster.graclus_cluster on empty inputs?

[{"_id":"66d0f10d5a1d13a447096764","body":"Thanks for reporting. Indeed, currently `graclus_cluster` assumes a well-defined graph. We could probably error out early in this we see empty inputs.","issue_id":1710340833904,"origin_id":1469815596,"user_origin_id":6945922,"create_time":1678878870,"update_time":1678878870,"id":1724969229692,"updated_at":"2024-08-29T22:07:09.691000Z","created_at":"2024-08-29T22:07:09.691000Z"},{"_id":"66d0f10d5a1d13a447096765","body":"This issue had no activity for **6 months**. It will be closed in **2 weeks** unless there is some new activity. Is this issue already resolved?","issue_id":1710340833904,"origin_id":1714805357,"user_origin_id":41898282,"create_time":1694479840,"update_time":1694479840,"id":1724969229695,"updated_at":"2024-08-29T22:07:09.695000Z","created_at":"2024-08-29T22:07:09.695000Z"}]

When calling `graclus_cluster(row, col, weight, numnodes)` on empty input, the call hangs. In particular by "empty inputs" I mean inputs like: `row=torch.Tensor([])`, `col=torch.Tensor([])`, `weight=torch.Tensor([])`, `numnodes=some positive integer`. While the call...

ssrothman

`radius_graph` bug: it might produce `max_num_neighbors + 1` neighbors for both CPU and GPU versions

[]

Hi Team, I found that `radius_graph` might produce `max_num_neighbors + 1` neighbors for both CPU and GPU versions. A minimal reproducer for CPU, adapted from the example in README: ```python...

stslxg-nv

Add num_points to fps as an alternative to ratio

[]

The current implementation of farthest point sampling only allows to set a ratio value to define the number of output points. Some use-cases require a fixed number of output points....

LarsHaalck

ARM64 torch_cluster.radius bug

[]

When executing torch_cluster.radius on my MacBook Pro 16-inch(Nov 2023, M3 Pro) even a minimal working example gives an error message. Problem does NOT persist on Server Ubuntu 22.04.4. Minimum working...

JSSchmidt

Missing `omp.h` error

[]

When install torch-cluster using the command ```bash python3 -m pip install torch-cluster ``` I get the following error: ```bash Defaulting to user installation because normal site-packages is not writeable Collecting...

bd-fauna

Err: source code compile

[]

In conda ven, `python setup.py install` but have sub err apper!! > /lib/python3.12/site-packages/torch/include/torch/csrc/jit/api/module.h:409:113: error: use of deleted function ‘torch::jit::Module::Module(torch::jit::Module&&)’ 409 | bool return_module) // if true include root itself as...

chenbq18

Add Haversine distance metric

[]

The Haversine (great-circle) distance metric would be IMO a very useful addition. It'd allow for better clustering of points residing on a sphere (particularly for point / node configurations where...

mishooax