Updates related to changes in NeXus definitions, made during the NFDI-MatWerk project meeting in Frankfurt

Author: atomprobe-tc

compositionspace/autophase.py

@@ -87,12 +87,12 @@ def automated_phase_assignment(self):
         trg = f"/entry{self.config['entry_id']}/autophase/result"
         grp = h5w.create_group(trg)
         grp.attrs["NX_class"] = "NXdata"
-        grp.attrs["axes"] = "axis_feature_identifier"
-        grp.attrs["axis_feature_identifier_indices"] = np.uint64(0)
+        grp.attrs["axes"] = "axis_feature_indices"
+        grp.attrs["axis_feature_indices_indices"] = np.uint64(0)
         grp.attrs["signal"] = "axis_feature_importance"
         # further attributes, to render it a proper NeXus NXdata object
         dst = h5w.create_dataset(
-            f"{trg}/axis_feature_identifier",
+            f"{trg}/axis_feature_indices",
             compression="gzip",
             compression_opts=1,
             data=np.asarray(sorted_indices, APT_UINT),

compositionspace/clustering.py

@@ -61,7 +61,7 @@ def run(self):
         dst = h5w.create_dataset(f"{trg}/sequence_index", data=np.uint64(sequence_idx))
         trg = f"/entry{self.config['entry_id']}/clustering/ic_opt"
         grp = h5w.create_group(trg)
-        grp.attrs["NX_class"] = "NXobject"
+        grp.attrs["NX_class"] = "NXprocess"
         h5w.close()
 
         # n_ic_runs = sum(1 for grpnm in ic_results_group_names if grpnm.startswith("cluster_analysis"))

compositionspace/meshing.py

@@ -51,7 +51,7 @@ def meshes_normals(config_file_path, input_file_path, output_file_path):
     # and other parameter from yaml config file like shown in other *.py files
     Node_list_plot = []
     no_elements = config["voxelization/edge_length"]
-    dist_cut = config["meshing/dist_cut"]
+    dist_cut = config["meshing/distance_cut"]
     normal_end_length = config["meshing/normal_end_length"]
     nodes_edit_lst = []
     simp_plot_edit_lst = []

compositionspace/preparation.py

@@ -91,9 +91,9 @@ def write_init_results(self):
         )
         trg = f"/entry{self.config['entry_id']}/config"
         grp = h5w.create_group(trg)
-        grp.attrs["NX_class"] = "NXserialized"
+        grp.attrs["NX_class"] = "NXnote"
         dst = h5w.create_dataset(f"{trg}/type", data="file")
-        dst = h5w.create_dataset(f"{trg}/path", data=self.config["config_file_path"])
+        dst = h5w.create_dataset(f"{trg}/file_name", data=self.config["config_file_path"])
         dst = h5w.create_dataset(f"{trg}/checksum", data=get_sha256(self.config["config_file_path"]))
         dst = h5w.create_dataset(f"{trg}/algorithm", data="sha256")
 
@@ -237,12 +237,12 @@ def write_voxelization_grid_info(self):
         )
         dst.attrs["units"] = "nm"
         dst = h5w.create_dataset(f"{trg}/extent", data=self.extent)
-        identifier_offset = 0  # we count cells starting from this value
+        index_offset = 0  # we count cells starting from this value
         dst = h5w.create_dataset(
-            f"{trg}/identifier_offset", data=np.uint64(identifier_offset)
+            f"{trg}/index_offset", data=np.uint64(index_offset)
         )
 
-        voxel_id = identifier_offset
+        voxel_id = index_offset
         position = np.zeros([c, 3], np.float64)
         for k in np.arange(0, self.extent[2]):
             z = self.aabb3d[2, 0] + (0.5 + k) * dedge
@@ -258,7 +258,7 @@ def write_voxelization_grid_info(self):
         dst.attrs["units"] = "nm"
         del position
 
-        voxel_id = identifier_offset
+        voxel_id = index_offset
         coordinate = np.zeros([c, 3], np.uint64)
         for k in np.arange(0, self.extent[2]):
             for j in np.arange(0, self.extent[1]):
@@ -295,7 +295,7 @@ def write_voxelization_results(self):
         h5w = h5py.File(self.config["results_file_path"], "a")
         trg = f"/entry{self.config['entry_id']}/voxelization/cg_grid"
         dst = h5w.create_dataset(
-            f"{trg}/voxel_identifier",
+            f"{trg}/indices_voxel",
             compression="gzip",
             compression_opts=1,
             data=self.voxel_identifier,
@@ -345,14 +345,15 @@ def write_voxelization_results(self):
         )
 
         # specimen group
-        trg = f"/entry{self.config['entry_id']}/specimen"
-        grp = h5w.create_group(f"{trg}")
-        grp.attrs["NX_class"] = "NXsample"
-        if str(self.config["specimen/type"]) in ["experiment", "simulation"]:
-            dst = h5w.create_dataset(f"{trg}/type", data=str(self.config["specimen/type"]))
-        else:
-            dst = h5w.create_dataset(f"{trg}/type", data="unknown")
-        dst = h5w.create_dataset(f"{trg}/atom_types", data=str(", ".join(list(atom_types))))
+        if "specimen/is_simulation" in self.config:
+            trg = f"/entry{self.config['entry_id']}/specimen"
+            grp = h5w.create_group(f"{trg}")
+            grp.attrs["NX_class"] = "NXsample"
+            if self.config["specimen/is_simulation"] is True:
+                dst = h5w.create_dataset(f"{trg}/is_simulation", data=True)
+            else:
+                dst = h5w.create_dataset(f"{trg}/is_simulation", data=False)
+            dst = h5w.create_dataset(f"{trg}/atom_types", data=str(", ".join(list(atom_types))))
         h5w.close()
 
     def run(self, recon_file_path: str, range_file_path: str):

compositionspace/segmentation.py

@@ -110,15 +110,15 @@ def perform_bics_minimization_and_write_results(self):
         aics = []
         bics = []
         n_clusters_queue = list(
-            range(1, self.config["segmentation/n_max_ic_cluster"] + 1)
+            range(1, self.config["segmentation/ic_opt/n_max_ic_cluster"] + 1)
         )
         for n_bics_cluster in n_clusters_queue:
             X_train = None
             C_mod = None
             if self.config["autophase/use"]:
                 print("Using results with automated phase assignment")
                 with h5py.File(self.config["results_file_path"], "r") as h5r:
-                    trg = f"/entry{self.config['entry_id']}/autophase/result/axis_feature_identifier"
+                    trg = f"/entry{self.config['entry_id']}/autophase/result/axis_feature_indices"
                     if trg in h5r:
                         descending_indices = h5r[trg][:]
                         # print(descending_indices)
@@ -193,8 +193,8 @@ def perform_bics_minimization_and_write_results(self):
         axis_dim = np.asarray(
             np.linspace(
                 1,
-                self.config["segmentation/n_max_ic_cluster"],
-                num=self.config["segmentation/n_max_ic_cluster"],
+                self.config["segmentation/ic_opt/n_max_ic_cluster"],
+                num=self.config["segmentation/ic_opt/n_max_ic_cluster"],
                 endpoint=True,
             ),
             APT_UINT,

pyproject.toml

@@ -11,9 +11,8 @@ authors = [
 description = "APT analysis tools"
 readme = "README.md"
 license = { file = "LICENSE" }
-requires-python = ">=3.7"
+requires-python = ">=3.8"
 classifiers = [
-    "Programming Language :: Python :: 3.7",
     "Programming Language :: Python :: 3.8",
     "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
@@ -26,7 +25,7 @@ dependencies = [
     "h5py",
     "scikit-learn",
     "trimesh",
-    "ifes_apt_tc_data_modeling>=0.2.1",
+    "ifes_apt_tc_data_modeling>=0.2.2",
     "pyyaml",
     "flatdict",
     "lxml",

tests/experiment_params.yaml

@@ -1,5 +1,8 @@
+identifier_analysis:
+reconstruction:
+ranging:
 specimen:
-  type: experiment
+  is_simulation: false
 voxelization:
   edge_length: 1
 autophase:
@@ -8,8 +11,9 @@ autophase:
   trunc_species: 2
   random_forest_classifier:
 segmentation:
-  n_max_ic_cluster: 5
-  gaussian_mixture:
+  ic_opt:
+    n_max_ic_cluster: 5
+    gaussian_mixture:
 clustering:
   dbscan:
     eps: 3