Fix type error in DropPointsByClass

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v3.4.0
+    rev: v4.5.0
     hooks:
       # list of supported hooks: https://pre-commit.com/hooks.html
       - id: trailing-whitespace
@@ -12,26 +12,26 @@ repos:
 
   # python code formatting
   - repo: https://github.com/psf/black
-    rev: 20.8b1
+    rev: 24.2.0
     hooks:
       - id: black
         args: [--line-length, "99"]
 
   # python import sorting
   - repo: https://github.com/PyCQA/isort
-    rev: 5.8.0
+    rev: 5.13.2
     hooks:
       - id: isort
 
   # yaml formatting
   - repo: https://github.com/pre-commit/mirrors-prettier
-    rev: v2.3.0
+    rev: v4.0.0-alpha.8
     hooks:
       - id: prettier
         types: [yaml]
 
   # python code analysis
   - repo: https://github.com/PyCQA/flake8
-    rev: 3.9.2
+    rev: 7.0.0
     hooks:
       - id: flake8

CHANGELOG.md

@@ -1,4 +1,7 @@
 # CHANGELOG
+### 3.8.2
+- fix: type error in edge case when dropping points in transforms
+
 ### 3.8.1
 - fix: propagate input las format to output las (in particular epsg which comes either from input or config)
 

myria3d/pctl/transforms/transforms.py

@@ -36,6 +36,9 @@ def subsample_data(data, num_nodes, choice):
             continue
         elif torch.is_tensor(item) and item.size(0) == num_nodes and item.size(0) != 1:
             data[key] = item[choice]
+        elif isinstance(item, np.ndarray) and item.shape[0] == num_nodes and item.shape[0] != 1:
+            data[key] = item[choice]
+
     return data
 
 
@@ -234,7 +237,5 @@ def __call__(self, data):
         if points_to_drop.sum() > 0:
             points_to_keep = torch.logical_not(points_to_drop)
             data = subsample_data(data, num_nodes=data.num_nodes, choice=points_to_keep)
-            # Here we also subsample these idx since we do not need to interpolate these points back
-            if "idx_in_original_cloud" in data:
-                data.idx_in_original_cloud = data.idx_in_original_cloud[points_to_keep]
+
         return data

package_metadata.yaml

@@ -1,4 +1,4 @@
-__version__: "3.8.1"
+__version__: "3.8.2"
 __name__: "myria3d"
 __url__: "https://github.com/IGNF/myria3d"
 __description__: "Deep Learning for the Semantic Segmentation of Aerial Lidar Point Clouds"

tests/myria3d/models/test_model.py

@@ -0,0 +1,28 @@
+import hydra
+from pytorch_lightning import LightningDataModule
+from tests.conftest import make_default_hydra_cfg
+
+from myria3d.models.model import Model
+
+
+def test_model_get_batch_tensor_by_enumeration():
+    config = make_default_hydra_cfg(
+        overrides=[
+            "predict.src_las=tests/data/toy_dataset_src/862000_6652000.classified_toy_dataset.100mx100m.las",
+            "datamodule.epsg=2154",
+            "work_dir=./../../..",
+            "datamodule.subtile_width=1",
+            "datamodule.hdf5_file_path=null",
+        ]
+    )
+
+    datamodule: LightningDataModule = hydra.utils.instantiate(config.datamodule)
+    datamodule._set_predict_data(config.predict.src_las)
+
+    model = Model(
+        neural_net_class_name="PyGRandLANet",
+        neural_net_hparams=dict(num_features=2, num_classes=7),
+    )
+    for batch in datamodule.predict_dataloader():
+        # Check that no error is raised ("TypeError: object of type 'numpy.int64' has no len()")
+        _ = model._get_batch_tensor_by_enumeration(batch.idx_in_original_cloud)

tests/myria3d/data/test_transforms.py → tests/myria3d/pctl/transforms/test_transforms.py

@@ -1,9 +1,9 @@
 import numpy as np
 import pytest
-import torch_geometric
 import torch
+import torch_geometric
 
-from myria3d.pctl.transforms.transforms import TargetTransform, DropPointsByClass
+from myria3d.pctl.transforms.transforms import DropPointsByClass, TargetTransform
 
 
 def test_TargetTransform_with_valid_config():
@@ -12,10 +12,12 @@ def test_TargetTransform_with_valid_config():
     # 1 becomes 0, and 6 becomes 1.
     classification_dict = {1: "unclassified", 6: "building"}
     tt = TargetTransform(classification_preprocessing_dict, classification_dict)
-
     y = np.array([1, 1, 2, 2, 6, 6])
-    data = torch_geometric.data.Data(x=None, y=y)
-    assert np.array_equal(tt(data).y, np.array([0, 0, 0, 0, 1, 1]))
+    idx = np.arange(6)
+    data = torch_geometric.data.Data(x=None, y=y, idx_in_original_cloud=idx)
+    out_data = tt(data)
+    assert np.array_equal(out_data.y, np.array([0, 0, 0, 0, 1, 1]))
+    assert np.array_equal(out_data.idx_in_original_cloud, idx)
 
 
 def test_TargetTransform_throws_type_error_if_invalid_classification_dict():
@@ -34,11 +36,16 @@ def test_DropPointsByClass():
     # points with class 65 are droped.
     y = torch.Tensor([1, 65, 65, 2, 65])
     x = torch.rand((5, 3))
-    data = torch_geometric.data.Data(x=x, y=y)
+    idx = np.arange(5)  # Not a tensor
+    data = torch_geometric.data.Data(x=x, y=y, idx_in_original_cloud=idx)
     drop_transforms = DropPointsByClass()
     transformed_data = drop_transforms(data)
     assert torch.equal(transformed_data.y, torch.Tensor([1, 2]))
     assert transformed_data.x.size(0) == 2
+    print(type(transformed_data.idx_in_original_cloud))
+    assert isinstance(transformed_data.idx_in_original_cloud, np.ndarray)
+    assert transformed_data.idx_in_original_cloud.size == 2
+    assert np.all(transformed_data.idx_in_original_cloud == np.array([0, 3]))
 
     # No modification
     x = torch.rand((3, 3))
@@ -47,3 +54,15 @@ def test_DropPointsByClass():
     transformed_data = drop_transforms(data)
     assert torch.equal(data.x, transformed_data.x)
     assert torch.equal(data.y, transformed_data.y)
+
+    # Keep one point only
+    y = torch.Tensor([1, 65, 65, 65, 65])
+    x = torch.rand((5, 3))
+    idx = np.arange(5)  # Not a tensor
+    data = torch_geometric.data.Data(x=x, y=y, idx_in_original_cloud=idx)
+    transformed_data = drop_transforms(data)
+    assert torch.equal(transformed_data.y, torch.Tensor([1]))
+    assert transformed_data.x.size(0) == 1
+    assert isinstance(transformed_data.idx_in_original_cloud, np.ndarray)
+    assert transformed_data.idx_in_original_cloud.shape[0] == 1
+    assert np.all(transformed_data.idx_in_original_cloud == np.array([0]))