Is it possible to slice dpf.Model().metadata.meshed_region.node ?

Christophe Dang

Hello,
I need to get the Id's and coordinates of the nodes of a meshing made with Workbench.
I write something like

myModel = dpf.Model(r"anywhere\file.rst")
myMeshing = myModel.metadata.meshed_region
length = len(myMehsing.nodes)
nodesId = [myMeshing.nodes[i].id for i in range(length)]
nodesCoordinates = [myMeshing.nodes[i].coordinates for i in range(length)]

But as I have more than 1e6 nodes, this takes a some time (about 26 seconds); Acceptable, as long as you do it once in a while.
I tried to use the Python slicing, nodesId = myMeshing.nodes[:].id, but I get a “TypeError: 'slice' object cannot be interpreted as an integer”.

So, is there an elegant and efficient way to get all the Id's and coordinates at once, or is the for loop the only solution?

aksh3000

Hello @Christophe Dang

The for loop is unnecessary: DPF can retrieve all node IDs and coordinates in bulk through the mesh’s node scoping and coordinate field. Since mesh.nodes is a DPF collection rather than a standard Python list, it does not support Python slicing directly; first expose the bulk data as NumPy arrays, then apply the same slice or Boolean mask to both the ID and coordinate arrays to preserve their row-by-row correspondence. This should be significantly faster for meshes with millions of nodes because it avoids creating and accessing each DPF Node object individually; np.asarray() may also avoid an unnecessary memory copy, although the main performance gain comes from bulk DPF access. Remember that NumPy slicing operates on array positions, not node-ID values.

example:

from ansys.dpf import core as dpf
import numpy as np

------------------------------------------------------------1. Load the result file and obtain the mesh------------------------------------------------------------

src = r"D:#SR\2026\test\file.rst"
model = dpf.Model(src)
mesh = model.metadata.meshed_region

------------------------------------------------------------2. Retrieve IDs and coordinates in bulk------------------------------------------------------------

node_ids = np.asarray(mesh.nodes.scoping.ids)

Node coordinates: shape (number_of_nodes, 3)Columns correspond to X, Y, and Z.

node_coordinates = np.asarray(mesh.nodes.coordinates_field.data)

print("Number of nodes:", node_ids.size)
print("IDs shape:", node_ids.shape)
print("Coordinates shape:", node_coordinates.shape)

Verify that each coordinate row corresponds to one node ID.

if node_coordinates.shape != (node_ids.size, 3):
raise ValueError(
"Unexpected coordinate array shape: "
f"{node_coordinates.shape}"
)

------------------------------------------------------------3. Basic slicing examples------------------------------------------------------------First 10 node IDs and their coordinates.

first_10_ids = node_ids[:10]
first_10_coordinates = node_coordinates[:10, :]

print("\nFirst 10 nodes:")

for node_id, (x, y, z) in zip(first_10_ids, first_10_coordinates):
print(
f"Node {node_id}: "
f"X={x:.6f}, Y={y:.6f}, Z={z:.6f}"
)

pydpf_numpy.py