update code, test and version number

CHANGELOG.md

@@ -1,3 +1,6 @@
+# 1.7.4
+- Add possibility to remove points of somes classes in standardize
+
 # 1.7.3
 - Add method to get a point cloud origin
 

pdaltools/_version.py

@@ -1,4 +1,4 @@
-__version__ = "1.7.3"
+__version__ = "1.7.4"
 
 
 if __name__ == "__main__":

pdaltools/standardize_format.py

@@ -35,7 +35,7 @@
     offset_z=0,
     dataformat_id=6,  # No color by default
     a_srs="EPSG:2154",
-    remove_class=[],  # Save no extra_dims
+    remove_points_from_class=[],  # remove points from class
 )
 
 
@@ -48,7 +48,7 @@ def parse_args():
     )
     parser.add_argument("--projection", default="EPSG:2154", type=str, help="Projection, eg. EPSG:2154")
     parser.add_argument(
-        "--remove_class",
+        "--remove_points_from_class",
         default=[],
         nargs="*",
         type=str,
@@ -75,19 +75,18 @@ def get_writer_parameters(new_parameters: Dict) -> Dict:
     return params
 
 
-def remove_points_from_las(input_file: str, output_file: str, class_removed) -> None:
+def remove_points_from_class(input_file: str, output_file: str, class_removed) -> None:
     pipeline = pdal.Pipeline() | pdal.Reader.las(input_file)
     pipeline.execute()
     points = pipeline.arrays[0]
     input_dimensions = list(points.dtype.fields.keys())
     dim_class = input_dimensions.index("Classification")
 
-    indice_pnts_delete = []
-    for i in range(0, len(points)):
-        if points[i][dim_class] in class_removed:
-            indice_pnts_delete.append(i)
+    indice_pts_delete = [id for id in range(0, len(points)) if points[id][dim_class] in class_removed]
+    points_pruned = np.delete(points, indice_pts_delete)
 
-    points_pruned = np.delete(points, indice_pnts_delete)
+    if len(points_pruned) == 0:
+        raise Exception("All points removed !")
 
     params = get_writer_parameters_from_reader_metadata(pipeline.metadata)
     pipeline_end = pdal.Pipeline(arrays=[points_pruned])
@@ -138,7 +137,7 @@ def standardize(input_file: str, output_file: str, params_from_parser: Dict) ->
     if args.remove_class:
         filename = os.path.basename(args.input_file)
         with tempfile.NamedTemporaryFile(suffix=filename) as tmp:
-            remove_points_from_las(args.input_file, tmp.name, args.remove_class)
+            remove_points_from_class(args.input_file, tmp.name, args.remove_points_from_class)
             standardize(tmp.name, args.output_file, params_from_parser)
     else:
         standardize(args.input_file, args.output_file, params_from_parser)

test/test_las_remove_dimensions.py

@@ -13,16 +13,22 @@
 ini_las = os.path.join(INPUT_DIR, "test_data_77055_627760_LA93_IGN69.laz")
 added_dimensions = ["DIM_1", "DIM_2"]
 
-def get_points(input_las : str):
+
+def get_points(input_las: str):
     pipeline_read_ini = pdal.Pipeline() | pdal.Reader.las(input_las)
     pipeline_read_ini.execute()
     return pipeline_read_ini.arrays[0]
 
-def append_dimension(input_las : str, output_las : str):
+
+def append_dimension(input_las: str, output_las: str):
     pipeline = pdal.Pipeline()
     pipeline |= pdal.Reader.las(input_las)
     pipeline |= pdal.Filter.ferry(dimensions="=>" + ", =>".join(added_dimensions))
-    pipeline |= pdal.Writer.las(output_las, extra_dims="all", forward="all", )
+    pipeline |= pdal.Writer.las(
+        output_las,
+        extra_dims="all",
+        forward="all",
+    )
     pipeline.execute()
 
 
@@ -52,10 +58,10 @@ def test_remove_one_dimension():
             las_remove_dimensions.remove_dimensions_from_las(tmp_las.name, ["DIM_1"], tmp_las_rm.name)
             points_end = get_points(tmp_las_rm.name)
 
-            assert list(points_end.dtype.fields.keys()).index("DIM_2") >= 0# should still contains DIM_2
+            assert list(points_end.dtype.fields.keys()).index("DIM_2") >= 0  # should still contains DIM_2
 
             with pytest.raises(ValueError):
-                list(points_end.dtype.fields.keys()).index("DIM_1") # should not have DIM_1
+                list(points_end.dtype.fields.keys()).index("DIM_1")  # should not have DIM_1
 
             with pytest.raises(TypeError):
                 numpy.array_equal(points_ini, points_end)  # output data should not be the same

test/test_standardize_format.py

@@ -3,12 +3,13 @@
 import shutil
 import subprocess as sp
 import platform
+import json
 from test.utils import EXPECTED_DIMS_BY_DATAFORMAT, get_pdal_infos_summary
 
 import pdal
 import pytest
 
-from pdaltools.standardize_format import exec_las2las, rewrite_with_pdal, standardize, remove_points_from_las
+from pdaltools.standardize_format import exec_las2las, rewrite_with_pdal, standardize, remove_points_from_class
 
 TEST_PATH = os.path.dirname(os.path.abspath(__file__))
 TMP_PATH = os.path.join(TEST_PATH, "tmp")
@@ -117,30 +118,64 @@ def test_standardize_malformed_laz():
     standardize(input_file, output_file, MUTLIPLE_PARAMS[0])
     assert os.path.isfile(output_file)
 
-def test_remove_points_from_class():
+
+@pytest.mark.parametrize(
+    "classes_to_remove",
+    [
+        [2, 3],
+        [2, 3, 4],
+        [0, 1, 2, 3, 4, 5, 6],
+    ],
+)
+def test_remove_points_from_class(classes_to_remove):
     input_file = os.path.join(TEST_PATH, "data/classified_laz/test_data_77050_627755_LA93_IGN69.laz")
     output_file = os.path.join(TMP_PATH, "test_remove_points_from_class.laz")
 
-    def count_points_by_classes(las_file: str, classes) -> int:
-        pipeline = pdal.Pipeline() | pdal.Reader.las(las_file)
-        pipeline.execute()
-        points = pipeline.arrays[0]
-        input_dimensions = list(points.dtype.fields.keys())
-        dim_class = input_dimensions.index('Classification')
-        nbPts = 0
-        for pt in points:
-            if pt[dim_class] in classes:
-                nbPts +=1
-        return nbPts
-
-    classes_to_remove = [2, 3]
+    def get_pipeline_metadata(pipeline):
+        try:
+            metadata = json.loads(pipeline.metadata)
+        except TypeError:
+            d_metadata = json.dumps(pipeline.metadata)
+            metadata = json.loads(d_metadata)
+        return metadata
+
+    # count points of class not in classes_to_remove (get the point we should have in fine)
+    pipeline = pdal.Pipeline() | pdal.Reader.las(input_file)
+    where = ""
+    for cl in classes_to_remove:
+        where += "Classification != " + str(cl)
+        if cl != classes_to_remove[-1]:
+            where += " && "
+    pipeline |= pdal.Filter.stats(dimensions="Classification", enumerate="Classification", where=where)
+    pipeline.execute()
+    metadata = get_pipeline_metadata(pipeline)
+    statistic = metadata["metadata"]["filters.stats"]["statistic"]
+    nb_points_to_get = statistic[0]["count"]
 
-    remove_points_from_las(input_file, output_file, classes_to_remove)
+    try:
+        remove_points_from_class(input_file, output_file, classes_to_remove)
+    except Exception as error: # error because all points are removed
+        assert nb_points_to_get == 0
+        return
 
     assert os.path.isfile(output_file)
-    assert count_points_by_classes(input_file,  classes_to_remove) > 0
-    assert count_points_by_classes(output_file,  classes_to_remove) == 0
 
+    def get_statistics(las_file: str):
+        pipeline = pdal.Pipeline() | pdal.Reader.las(las_file)
+        pipeline |= pdal.Filter.stats(dimensions="Classification", enumerate="Classification")
+        pipeline.execute()
+        metadata = get_pipeline_metadata(pipeline)
+        statistic = metadata["metadata"]["filters.stats"]["statistic"]
+        return statistic[0]["count"], statistic[0]["values"]
+
+    nb_points_before, class_before = get_statistics(input_file)
+    nb_points_after, class_after = get_statistics(output_file)
+
+    assert nb_points_before > 0
+    assert nb_points_before > nb_points_after
+    assert set(classes_to_remove).issubset(set(class_before))
+    assert not set(classes_to_remove).issubset(set(class_after))
+    assert nb_points_after == nb_points_to_get
 
 
 if __name__ == "__main__":