Encounted UserWarning when using dataloader when I need to process some graph data and part of them have only one nodes

[wangyu-sd]

🐛 Describe the bug

Hello, when I attempted to utilize the torch_geometric.loader.DataLoader object to load my dataset, I encountered the following user warning:

/root/miniconda3/envs/chem/lib/python3.11/site-packages/torch_geometric/data/collate.py:204: UserWarning: An output with one or more elements was resized since it had shape [16314], which does not match the required output shape [2, 16314]. This behavior is deprecated, and in a future PyTorch release outputs will not be resized unless they have zero elements. You can explicitly reuse an out tensor t by resizing it, inplace, to zero elements with t.resize_(0). (Triggered internally at ../aten/src/ATen/native/Resize.cpp:28.)

value = torch.cat(values, dim=cat_dim or 0, out=out)

It suggested that the issue stemmed from edge_index, where the shape of the tensor out was initialized as [num_edges] instead of the expected [2, num_edges]. Upon reviewing the related source code located at
/root/miniconda3/envs/chem/lib/python3.11/site-packages/torch_geometric/data/collate.py:204, I discovered:

        out = None
        if torch.utils.data.get_worker_info() is not None:
            # Write directly into shared memory to avoid an extra copy:
            numel = sum(value.numel() for value in values)
            if torch_geometric.typing.WITH_PT20:
                storage = elem.untyped_storage()._new_shared(
                    numel * elem.element_size(), device=elem.device)
            elif torch_geometric.typing.WITH_PT112:
                storage = elem.storage()._new_shared(numel, device=elem.device)
            else:
                storage = elem.storage()._new_shared(numel)
            shape = list(elem.size())
            if cat_dim is None or elem.dim() == 0:
                shape = [len(values)] + shape
            else:
                shape[cat_dim] = int(slices[-1])
            out = elem.new(storage).resize_(*shape)

        value = torch.cat(values, dim=cat_dim or 0, out=out)

I propose the reason might be as follows: when the key is assigned the value 'edge_index' the 'cat_dim' is designated as '-1'. In scenarios where a graph comprising solely one node is presented as the initial element in the
'values' array, the shape of elem is determined to be [2,] instead of [2, num_edges_0]. This discrepancy leads to the ultimate shape of out being incorrectly set to '[num_edges_all]' following the execution of the command 'shape[-1] = int(slices[-1])'.

Are there any solutions to address this particular situation?

Versions

PyTorch version: 2.2.2+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.3 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version: Could not collect
CMake version: version 3.27.4
Libc version: glibc-2.35

Python version: 3.11.0 | packaged by conda-forge | (main, Jan 14 2023, 12:27:40) [GCC 11.3.0] (64-bit runtime)
Python platform: Linux-5.4.0-100-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA GeForce RTX 3090
GPU 1: NVIDIA GeForce RTX 3090
GPU 2: NVIDIA GeForce RTX 3090
GPU 3: NVIDIA GeForce RTX 3090
GPU 4: NVIDIA GeForce RTX 3090
GPU 5: NVIDIA GeForce RTX 3090
GPU 6: NVIDIA GeForce RTX 3090
GPU 7: NVIDIA GeForce RTX 3090

Nvidia driver version: 550.54.14
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.8.9.5
/usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.9.5
/usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.9.5
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.9.5
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.9.5
/usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.9.5
/usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.9.5
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True

CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 46 bits physical, 48 bits virtual
Byte Order: Little Endian
CPU(s): 80
On-line CPU(s) list: 0-79
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Gold 6133 CPU @ 2.50GHz
CPU family: 6
Model: 85
Thread(s) per core: 2
Core(s) per socket: 20
Socket(s): 2
Stepping: 4
CPU max MHz: 3000.0000
CPU min MHz: 1000.0000
BogoMIPS: 5000.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cdp_l3 invpcid_single pti intel_ppin ssbd mba ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm mpx rdt_a avx512f avx512dq rdseed adx smap clflushopt clwb intel_pt avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local dtherm ida arat pln pts pku ospke md_clear flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 1.3 MiB (40 instances)
L1i cache: 1.3 MiB (40 instances)
L2 cache: 40 MiB (40 instances)
L3 cache: 55 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-19,40-59
NUMA node1 CPU(s): 20-39,60-79
Vulnerability Itlb multihit: KVM: Mitigation: Split huge pages
Vulnerability L1tf: Mitigation; PTE Inversion; VMX conditional cache flushes, SMT vulnerable
Vulnerability Mds: Mitigation; Clear CPU buffers; SMT vulnerable
Vulnerability Meltdown: Mitigation; PTI
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Full generic retpoline, IBPB conditional, IBRS_FW, STIBP conditional, RSB filling
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Mitigation; Clear CPU buffers; SMT vulnerable

Versions of relevant libraries:
[pip3] numpy==1.26.4
[pip3] pytorch-lightning==2.2.2
[pip3] torch==2.2.2
[pip3] torch_cluster==1.6.3+pt22cu121
[pip3] torch_geometric==2.5.3
[pip3] torch_scatter==2.1.2+pt22cu121
[pip3] torch_sparse==0.6.18+pt22cu121
[pip3] torch_spline_conv==1.2.2+pt22cu121
[pip3] torchaudio==2.2.2
[pip3] torchdata==0.7.1
[pip3] torchmetrics==1.3.2
[pip3] torchvision==0.17.2
[pip3] triton==2.2.0
[conda] numpy 1.26.4 py311h24aa872_0
[conda] numpy-base 1.26.4 py311hbfb1bba_0
[conda] pytorch-lightning 2.2.2 pypi_0 pypi
[conda] torch 2.2.2 pypi_0 pypi
[conda] torch-cluster 1.6.3+pt22cu121 pypi_0 pypi
[conda] torch-geometric 2.5.3 pypi_0 pypi
[conda] torch-scatter 2.1.2+pt22cu121 pypi_0 pypi
[conda] torch-sparse 0.6.18+pt22cu121 pypi_0 pypi
[conda] torch-spline-conv 1.2.2+pt22cu121 pypi_0 pypi
[conda] torchaudio 2.2.2 pypi_0 pypi
[conda] torchdata 0.7.1 pypi_0 pypi
[conda] torchmetrics 1.3.2 pypi_0 pypi
[conda] torchvision 0.17.2 pypi_0 pypi
[conda] triton 2.2.0 pypi_0 pypi

[rusty1s]

Do you mind clarifying why edge_index would be of shape [2] in your case? Ideally, it should be [2, 0], e.g., via

data.edge_index = torch.empty(2, 0, dtype=torch.long)