============================= test session starts ==============================
platform linux -- Python 3.8.0, pytest-7.1.2, pluggy-0.13.1
rootdir: /mnt/d/gitlab/shape-analysis-github
collected 12 items

tests/test_architectures.py ..FF                                         [ 33%]
tests/test_build_sparse_adjacency.py .F.                                 [ 58%]
tests/test_compute_face_area_and_angle.py .                              [ 66%]
tests/test_metric_conv.py ....                                           [100%]

=================================== FAILURES ===================================
_________________________ test_metric_conv_net[False] __________________________

classification = False

    @pytest.mark.parametrize("classification", [True, False])
    def test_metric_conv_net(classification: bool):
        """
        Test of ``MetricConvNet`` from the ``metric_conv`` mdoule.
    
        :param classification: Boolean indicating whether to convert model to classification network
        """
        kwargs["classification"] = classification
        model = architectures.MetricConvNet(in_feats, out_feats, **kwargs)
>       out = model(features_per_vertex, pos, edges, faces)

tests/test_architectures.py:65: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
/home/rtrachel/miniconda3/envs/metric-conv/lib/python3.8/site-packages/torch/nn/modules/module.py:1110: in _call_impl
    return forward_call(*input, **kwargs)
models/architectures.py:358: in forward
    out = self.conv6(x, vertices, edges, faces)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = MetricConvNet(
  (convs): ModuleList(
    (0): MetricConv(
      (metric): FaceMetric(
        (metric_mlp): MetricMLP...ear(in_features=32, out_features=18, bias=True)
          )
        )
      )
    )
  )
  (nonlinear): ELU(alpha=1.0)
)
name = 'conv6'

    def __getattr__(self, name: str) -> Union[Tensor, 'Module']:
        if '_parameters' in self.__dict__:
            _parameters = self.__dict__['_parameters']
            if name in _parameters:
                return _parameters[name]
        if '_buffers' in self.__dict__:
            _buffers = self.__dict__['_buffers']
            if name in _buffers:
                return _buffers[name]
        if '_modules' in self.__dict__:
            modules = self.__dict__['_modules']
            if name in modules:
                return modules[name]
>       raise AttributeError("'{}' object has no attribute '{}'".format(
            type(self).__name__, name))
E       AttributeError: 'MetricConvNet' object has no attribute 'conv6'

/home/rtrachel/miniconda3/envs/metric-conv/lib/python3.8/site-packages/torch/nn/modules/module.py:1185: AttributeError
____________________________ test_linear_metric_net ____________________________

    def test_linear_metric_net():
        """
        Test of ``LinearMetricNet`` from the ``metric_conv`` module.
        """
        model = architectures.LinearMetricNet(in_feats, out_feats, **kwargs)
>       out = model(pos, edges, faces)

tests/test_architectures.py:78: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = LinearMetricNet(
  (fc0): Linear(in_features=3, out_features=16, bias=True)
  (conv1): MetricConv(
    (metric): FaceM...1): Linear(in_features=128, out_features=256, bias=True)
  (fc2): Linear(in_features=256, out_features=86, bias=True)
)
input = (tensor([[ 0.0000,  0.0000,  0.2358],
        [ 0.0000,  0.7468,  0.0000],
        [ 0.5387,  0.0000,  0.0000],
      ...  [1, 2, 3, 2, 3, 3, 0, 0, 0, 1, 1, 2]]), tensor([[0, 1, 2],
        [0, 2, 3],
        [0, 1, 3],
        [1, 3, 2]]))
kwargs = {}
forward_call = <bound method LinearMetricNet.forward of LinearMetricNet(
  (fc0): Linear(in_features=3, out_features=16, bias=True)
 ...): Linear(in_features=128, out_features=256, bias=True)
  (fc2): Linear(in_features=256, out_features=86, bias=True)
)>

    def _call_impl(self, *input, **kwargs):
        forward_call = (self._slow_forward if torch._C._get_tracing_state() else self.forward)
        # If we don't have any hooks, we want to skip the rest of the logic in
        # this function, and just call forward.
        if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
                or _global_forward_hooks or _global_forward_pre_hooks):
>           return forward_call(*input, **kwargs)
E           TypeError: forward() missing 1 required positional argument: 'faces'

/home/rtrachel/miniconda3/envs/metric-conv/lib/python3.8/site-packages/torch/nn/modules/module.py:1110: TypeError
________ TestBuildSparseAdjacency.test_build_sparse_adjacency_symmetric ________

self = <test_build_sparse_adjacency.TestBuildSparseAdjacency object at 0x7fcfac7990d0>

    def test_build_sparse_adjacency_symmetric(self):
        """
        Test for `metric.build_sparse_adjacency` when `symmetric=True`.
        """
        # Build a dense sparse matrix of size (n_nodes,n_nodes) with nnz non-zero elements.
        n_nodes = 10
        nnz = n_nodes
    
        # Non-zero entries
        idx1 = []
        idx2 = []
        weights1 = []
        weights2 = []
        weights_diag = torch.rand(n_nodes)
    
        matrix = torch.diag(1 / (1 + weights_diag ** 2))
        for _ in range(nnz):
            i = torch.randint(0, n_nodes, (1,))
            j = torch.randint(0, n_nodes, (1,))
            if [i, j] in idx1 or [j, i] in idx1 or i == j:
                # Make sure that an entry will not have multiple values
                continue
            w1 = torch.rand(1)
            w2 = torch.rand(1)
            matrix[i, j] = 1 / (1 + w1 ** 2)
            matrix[j, i] = 1 / (1 + w2 ** 2)
    
            idx1.append([i, j])
            weights1.append(w1)
    
            idx2.append([j, i])
            weights2.append(w2)
    
        idx1 = torch.tensor(idx1).t()
        idx2 = torch.tensor(idx2).t()
        weights1 = torch.tensor(weights1)
        weights2 = torch.tensor(weights2)
    
        sum_row = matrix.sum(dim=1)
        matrix = matrix / sum_row.view(n_nodes, 1)
        answer = 0.5 * (matrix + matrix.t())  # Symmetrize matrix (this is what will be tested against)
    
        idx_identity = torch.cat(
            (
                torch.arange(n_nodes).view(1, n_nodes),
                torch.arange(n_nodes).view(1, n_nodes),
            ),
            dim=0,
        )
    
        idx = torch.cat((idx_identity, idx1, idx2), dim=1)
        weights = torch.cat((weights_diag, weights1, weights2))
        x = metric.build_sparse_adjacency(idx, weights, device="cpu", symmetric=True, remove_reference=False)
>       assert (torch.abs(x.to_dense() - answer) < 1e-6).sum() == answer.nelement()
E       assert tensor(71) == 100
E        +  where tensor(71) = <built-in method sum of Tensor object at 0x7fcfac68f680>()
E        +    where <built-in method sum of Tensor object at 0x7fcfac68f680> = tensor([[0.0193, 0.0000, 0.0000, 0.0261, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n         0.0559],\n        [0.0000, 0....,\n         0.0000],\n        [0.0378, 0.0000, 0.0000, 0.0000, 0.0079, 0.0210, 0.0000, 0.0000, 0.0000,\n         0.0035]]) < 1e-06.sum
E        +      where tensor([[0.0193, 0.0000, 0.0000, 0.0261, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n         0.0559],\n        [0.0000, 0....,\n         0.0000],\n        [0.0378, 0.0000, 0.0000, 0.0000, 0.0079, 0.0210, 0.0000, 0.0000, 0.0000,\n         0.0035]]) = <built-in method abs of type object at 0x7fcfbfd4bc00>((tensor([[0.3171, 0.0000, 0.0000, 0.3176, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n         0.3654],\n        [0.0000, 0....,\n         0.0000],\n        [0.2717, 0.0000, 0.0000, 0.0000, 0.2000, 0.2735, 0.0000, 0.0000, 0.0000,\n         0.2548]]) - tensor([[0.2978, 0.0000, 0.0000, 0.2915, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n         0.3094],\n        [0.0000, 0....,\n         0.0000],\n        [0.3094, 0.0000, 0.0000, 0.0000, 0.2079, 0.2945, 0.0000, 0.0000, 0.0000,\n         0.2513]])))
E        +        where <built-in method abs of type object at 0x7fcfbfd4bc00> = torch.abs
E        +        and   tensor([[0.3171, 0.0000, 0.0000, 0.3176, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n         0.3654],\n        [0.0000, 0....,\n         0.0000],\n        [0.2717, 0.0000, 0.0000, 0.0000, 0.2000, 0.2735, 0.0000, 0.0000, 0.0000,\n         0.2548]]) = <built-in method to_dense of Tensor object at 0x7fcfac68fea0>()
E        +          where <built-in method to_dense of Tensor object at 0x7fcfac68fea0> = tensor(indices=tensor([[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 0, 5, 4, 6, 1, 5, 5, 4,\n                        4, 9, 3, 6, 9...\n                      0.1448, 0.1200, 0.1315, 0.1559, 0.0986]),\n       size=(10, 10), nnz=40, layout=torch.sparse_coo).to_dense
E        +  and   100 = <built-in method nelement of Tensor object at 0x7fcfac793720>()
E        +    where <built-in method nelement of Tensor object at 0x7fcfac793720> = tensor([[0.2978, 0.0000, 0.0000, 0.2915, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000,\n         0.3094],\n        [0.0000, 0....,\n         0.0000],\n        [0.3094, 0.0000, 0.0000, 0.0000, 0.2079, 0.2945, 0.0000, 0.0000, 0.0000,\n         0.2513]]).nelement

tests/test_build_sparse_adjacency.py:119: AssertionError
=========================== short test summary info ============================
FAILED tests/test_architectures.py::test_metric_conv_net[False] - AttributeEr...
FAILED tests/test_architectures.py::test_linear_metric_net - TypeError: forwa...
FAILED tests/test_build_sparse_adjacency.py::TestBuildSparseAdjacency::test_build_sparse_adjacency_symmetric
========================= 3 failed, 9 passed in 1.37s ==========================