Heightfield support #1189#1547
Merged
adenzler-nvidia merged 30 commits intoFeb 16, 2026
Merged
Conversation
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Contributor
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
📝 WalkthroughWalkthroughAdds first-class heightfield support: new Heightfield type, loader utilities, MJCF import/parse and ModelBuilder API, geometry inertia/radius handling, MuJoCo exporter integration, and comprehensive tests. Changes
Sequence DiagramsequenceDiagram
participant App as Application
participant Builder as ModelBuilder
participant Geometry as Geometry Module
participant Importer as MJCF Importer
participant Solver as SolverMuJoCo
participant MuJoCo as MuJoCo Engine
App->>Builder: add_shape_heightfield(body, heightfield, ...)
Builder->>Geometry: register shape (GeoType.HFIELD, src=Heightfield)
Geometry->>Geometry: compute_shape_inertia(src, scale)
Geometry->>Geometry: compute_shape_radius(src, scale)
Builder-->>App: shape created
App->>Importer: import MJCF (may contain <hfield> assets)
Importer->>Importer: parse hfield asset (file or inline)
Importer->>Geometry: construct Heightfield(data, nrow,ncol,size)
Importer->>Builder: add_shape_heightfield(...)
App->>Solver: export model to MuJoCo
Solver->>Geometry: fetch heightfield from shape_source
Geometry-->>Solver: {nrow,ncol,size,data}
Solver->>MuJoCo: add_hfield(nrow,ncol,size,data)
Solver->>MuJoCo: create geom referencing hfield
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related PRs
Suggested reviewers
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches🧪 Generate unit tests (beta)
Tip Issue Planner is now in beta. Read the docs and try it out! Share your feedback on Discord. Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Codecov Report❌ Patch coverage is 📢 Thoughts on this report? Let us know! |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 6
🤖 Fix all issues with AI agents
In `@newton/_src/geometry/types.py`:
- Around line 474-485: The heightfield inertia calculation in
compute_shape_inertia (when compute_inertia is True) currently passes
half-extents size_x and size_y directly to compute_box_inertia; update the call
so the horizontal extents are converted to full extents by multiplying size_x
and size_y by 2.0 (i.e., pass size_x * 2.0 and size_y * 2.0), keeping
size_z/avg_height and size_base logic unchanged; refer to symbols
compute_inertia, size_x, size_y, avg_height, self._data, and compute_box_inertia
to locate and modify the call.
- Around line 502-516: The return type annotation on finalize is incorrect:
change the annotated return type from wp.uint64 to int and update the docstring
return type description accordingly; locate the finalize method (uses
self.warp_array = wp.array(...), returns self.warp_array.ptr) and adjust its
signature and docstring to indicate it returns a Python int (consistent with
Mesh.finalize and usage in wp.array(geo_sources, dtype=wp.uint64)).
In `@newton/_src/solvers/mujoco/solver_mujoco.py`:
- Around line 3081-3087: The warning branch in the GeoType.HFIELD handling uses
an undefined name `verbose`, causing a NameError; update the check to use the
file's existing verbosity flag (the same variable used elsewhere in this module,
e.g. self.verbose or verbosity) so the code reads something like "if <existing
verbosity flag>:" before printing the warning about `hfield_src` being None,
keeping the rest of the logic (skip the shape) unchanged; ensure you update the
reference in the block that inspects `model.shape_source[shape]` and mentions
`shape` and `hfield_src`.
In `@newton/_src/utils/heightfield.py`:
- Around line 90-95: Validate the binary header and payload size before
reshaping: after reading header = np.fromfile(f, dtype=np.int32, count=2) ensure
header has length 2 and both values are positive integers, compute
expected_count = int(header[0]) * int(header[1]), then check that data =
np.fromfile(f, dtype=np.float32, count=expected_count) returned exactly
expected_count elements (and raise a ValueError with a clear message including
the header values and expected vs actual counts if not); update the logic around
the file-read/reshape in heightfield.py (the header, data, filename variables)
to fail fast when header is malformed or payload size mismatches instead of
silently reshaping.
In `@newton/_src/utils/import_mjcf.py`:
- Around line 313-332: The parsed size tuple from np.fromstring (size_str →
size) can be shorter than 4 and lead to malformed Heightfield entries; update
the parsing in import_mjcf.py to validate and normalize size to exactly 4
floats: parse with np.fromstring, then if the resulting array has fewer than 4
elements, pad it (e.g., with 0.0 for missing entries) and if longer, truncate to
4, and finally convert to a tuple before storing in hfield_assets under the
"size" key so downstream consumers expecting (size_x, size_y, size_z, size_base)
always receive four values.
- Around line 641-665: The heightfield size is not being multiplied by the
global scale, causing mismatched extents; update the hfield handling so the size
passed from hfield_asset["size"] is scaled (e.g., multiply by the existing scale
variable) before creating the Heightfield via load_heightfield_from_file or
before passing into builder.add_shape_heightfield; locate the hfield branch that
uses geom_type == "hfield", hfield_assets, load_heightfield_from_file and
builder.add_shape_heightfield and apply scale to hfield_asset["size"] (and any
related dimensions like nrow/ncol if needed) so the heightfield respects
parse_mjcf's scale parameter.
🧹 Nitpick comments (7)
newton/_src/geometry/types.py (2)
486-489: Intentional divergence fromMeshdefaults —mass=0.0vsmass=1.0.When
compute_inertia=False,Meshdefaults tomass=1.0whileHeightfielddefaults tomass=0.0. This is reasonable since heightfields are typically static terrain, but the identity inertia tensor (wp.mat33(np.eye(3))) is meaningless for a zero-mass body. Consider usingwp.mat33()(zero matrix) instead, which is whatcompute_shape_inertiareturns for the static path anyway. This would make the intent clearer.Proposed fix
else: - self.I = wp.mat33(np.eye(3)) + self.I = wp.mat33() self.mass = 0.0 # Static by default self.com = wp.vec3()
529-538: Hashing large heightfields viatuple(self._data.flatten())can be very slow.For a 1024×1024 heightfield (~1M values), converting to a tuple and hashing will be expensive. This is the same pattern used by
Mesh.__hash__, so it's consistent, but heightfield grids can be significantly larger than typical meshes. Consider using a faster hash approach (e.g.,hash(self._data.tobytes())) if performance becomes an issue.newton/_src/geometry/inertia.py (2)
535-557: Inertia scaling logic is duplicated with the MESH/SDF branch (lines 560-579).Lines 540-554 are nearly identical to lines 561-579. Consider extracting the shared scaling logic into a small helper to reduce duplication. However, since this mirrors the existing pattern and the HFIELD branch intentionally skips the
is_solidcheck, keeping it separate is acceptable for now.
538-538: Missingis_solidconsistency check unlike the MESH branch.The MESH branch at line 560 checks
src.is_solid == is_solid, but the HFIELD branch omits this. If this is intentional (heightfields are always treated as solid), consider adding a brief comment to document the decision.newton/_src/geometry/utils.py (1)
85-85: Type hint forsrcparameter doesn't includeHeightfield.The function signature is
src: Mesh | SDF | Nonebut now handlesHeightfieldobjects via theGeoType.HFIELDbranch. Update the type hint to includeHeightfield.Proposed fix
Add the import at the top of the file alongside other geometry types:
from .types import ( SDF, GeoType, + Heightfield, Mesh, )Then update the function signature:
-def compute_shape_radius(geo_type: int, scale: Vec3, src: Mesh | SDF | None) -> float: +def compute_shape_radius(geo_type: int, scale: Vec3, src: Mesh | SDF | Heightfield | None) -> float:newton/_src/utils/import_mjcf.py (1)
641-641: Reusingparse_meshesto gate heightfield parsing is semantically misleading.A user disabling
parse_meshes=Falseto skip mesh loading would also inadvertently skip heightfield loading with no indication. Consider introducing a dedicatedparse_heightfieldsparameter, or at minimum document the current coupling in theparse_meshesdocstring.newton/tests/test_heightfield.py (1)
142-181: Optional: clean up the temporary binary file.This avoids temp-file buildup across repeated test runs (especially locally).
Proposed fix
+import os ... with tempfile.NamedTemporaryFile( suffix=".bin", delete=False, ) as f: tmp_path = f.name np.array([nrow, ncol], dtype=np.int32).tofile(f) elevation.tofile(f) + try: ... - builder = newton.ModelBuilder() - builder.add_mjcf( - mjcf, - parse_meshes=True, - path_resolver=resolver, - ) + builder = newton.ModelBuilder() + builder.add_mjcf( + mjcf, + parse_meshes=True, + path_resolver=resolver, + ) + finally: + os.unlink(tmp_path)
Member
|
Thanks @nvtw could you advise Viktor on how to potentially unify with the heightfields you are partially supporting in the unified collision pipeline? This being MuJoCo-only for now is ok, but we should make sure at least the public API can handle proper Newton heightfields. |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 5
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
newton/_src/geometry/utils.py (1)
85-85: 🛠️ Refactor suggestion | 🟠 MajorType hint for
srcshould includeHeightfield.The function now accesses
src.sizewhengeo_type == GeoType.HFIELD, but the type annotation isMesh | SDF | None. This will cause type-checker errors and is misleading for callers.Proposed fix
Add the import at the top of the file (or use a
TYPE_CHECKINGguard) and update the signature:from .types import ( SDF, GeoType, + Heightfield, Mesh, )-def compute_shape_radius(geo_type: int, scale: Vec3, src: Mesh | SDF | None) -> float: +def compute_shape_radius(geo_type: int, scale: Vec3, src: Mesh | SDF | Heightfield | None) -> float:newton/_src/geometry/inertia.py (1)
458-461: 🛠️ Refactor suggestion | 🟠 MajorType hint for
srcshould includeHeightfield.Same as in
utils.py— the new HFIELD branch accessessrc.has_inertia,src.mass,src.com,src.I, which areHeightfieldattributes, but the signature only allowsSDF | Mesh | None.Proposed fix
+from .types import ( + SDF, + GeoType, + Heightfield, + Mesh, + Vec3, +)def compute_shape_inertia( type: int, scale: Vec3, - src: SDF | Mesh | None, + src: SDF | Mesh | Heightfield | None, density: float, is_solid: bool = True, thickness: list[float] | float = 0.001, ) -> tuple[float, wp.vec3, wp.mat33]:
🤖 Fix all issues with AI agents
In `@newton/_src/geometry/utils.py`:
- Around line 107-119: The HFIELD bounding-sphere code currently ignores the
fourth element (size_base) and only uses size_z for vertical extent; unpack the
fourth value from src.size (e.g., size_x, size_y, size_z, size_base) and compute
the vertical extent as the larger of size_z and size_base (vertical_extent =
max(size_z, size_base)), then apply scale[2] to that extent (max_z =
vertical_extent * scale[2]) before computing the radius using half_x, half_y and
max_z so the sphere covers downward extensions as well.
In `@newton/_src/solvers/mujoco/solver_mujoco.py`:
- Around line 3081-3087: The branch handling GeoType.HFIELD uses an undefined
variable verbose when hfield_src is None; change the check to use
wp.config.verbose (consistent with nearby code) or call a warning logger
instead. Locate the block around GeoType.HFIELD in solver_mujoco.py where
hfield_src = model.shape_source[shape] and replace references to verbose with
wp.config.verbose (or add a warning via processLogger/warnings) so the missing
heightfield source is skipped without raising a NameError.
In `@newton/_src/utils/heightfield.py`:
- Around line 64-95: In _load_elevation_data, validate the binary file header
and payload against the expected nrow and ncol: after reading header =
np.fromfile(..., count=2), check that header has two entries and that
header[0]==nrow and header[1]==ncol (raise ValueError if they differ), then read
the payload and verify its length equals nrow*ncol (raise ValueError if payload
size is wrong) before calling reshape; this ensures the binary path enforces the
same dimension checks as the PNG path and prevents silent reordering when using
header values.
In `@newton/_src/utils/import_mjcf.py`:
- Around line 313-332: The code currently ignores an inline elevation attribute
on <hfield>; detect hfield.attrib.get("elevation") and handle it: if elevation
is present and no file_attr, parse it with np.fromstring(..., sep=" ",
dtype=np.float32), verify its length equals nrow*ncol (reshape to (nrow, ncol)),
and store it in the hfield_assets entry under the key "elevation" so downstream
code (instead of calling load_heightfield_from_file(None,...)) can use it; if
both elevation and file are present, emit a clear warning (use the module's
logger or warnings.warn) that the file takes precedence (or vice versa) and do
not silently drop elevation; if the parsed elevation length mismatches
nrow*ncol, log/warn and fall back to existing behavior. Ensure you reference the
existing variables and structures (hfield.attrib, nrow, ncol, file_attr,
file_path, size_str, hfield_assets, and load_heightfield_from_file) when
implementing the change.
In `@newton/tests/test_heightfield.py`:
- Around line 136-181: The test_mjcf_hfield_binary_file leaves the temp binary
file behind because NamedTemporaryFile was created with delete=False; wrap the
temp file creation and the subsequent usage (resolver, builder.add_mjcf and
assertions) in a try/finally and in the finally call os.remove(tmp_path) (or
pathlib.Path(tmp_path).unlink()) to ensure the file is deleted; reference
tmp_path and the resolver closure that returns it so cleanup always runs even if
builder.add_mjcf or assertions raise.
🧹 Nitpick comments (2)
newton/_src/geometry/inertia.py (1)
535-557: Duplicated inertia-scaling logic — consider extracting a helper.Lines 540–554 are an exact copy of the mesh inertia-scaling path at lines 561–578 (same
mass_ratio,c_new,I_newformulas). This violates DRY and means any future fix to the scaling math needs to be applied in two places.Consider extracting a small helper, e.g.:
Sketch
def _scale_precomputed_inertia(src, scale, density): """Scale precomputed mass/com/inertia from a geometry source.""" m, c, I = src.mass, src.com, src.I scale = wp.vec3(scale) sx, sy, sz = scale mass_ratio = sx * sy * sz * density m_new = m * mass_ratio c_new = wp.cw_mul(c, scale) Ixx = I[0, 0] * (sy**2 + sz**2) / 2 * mass_ratio Iyy = I[1, 1] * (sx**2 + sz**2) / 2 * mass_ratio Izz = I[2, 2] * (sx**2 + sy**2) / 2 * mass_ratio Ixy = I[0, 1] * sx * sy * mass_ratio Ixz = I[0, 2] * sx * sz * mass_ratio Iyz = I[1, 2] * sy * sz * mass_ratio I_new = wp.mat33([[Ixx, Ixy, Ixz], [Ixy, Iyy, Iyz], [Ixz, Iyz, Izz]]) return m_new, c_new, I_newThen both the HFIELD and MESH/SDF/CONVEX_MESH branches can call
_scale_precomputed_inertia(src, scale, density).newton/_src/utils/import_mjcf.py (1)
641-641: Overloadingparse_meshesto also gate heightfield parsing.
parse_meshesis documented as controlling only"mesh"geometries (line 180: "Whether geometries of type"mesh"should be parsed"). Gating heightfields behind this same flag meansparse_meshes=Falsesilently skips heightfields too, which may surprise callers.Consider either:
- Adding a dedicated
parse_heightfieldsparameter, or- At minimum, updating the
parse_meshesdocstring to clarify it also controls heightfields.
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@newton/_src/solvers/mujoco/solver_mujoco.py`:
- Around line 3097-3103: Replace the incorrect call to spec.add_hfield(...)
(which passes keyword args) with creating the hfield object and setting its
fields directly: call spec.add_hfield() with no arguments to get an mjsHField,
then assign hfield.name = hfield_name, hfield.nrow = nrow, hfield.ncol = ncol,
hfield.size = size, and hfield.userdata = elevation_data so the MuJoCo API is
used correctly.
🧹 Nitpick comments (4)
newton/_src/solvers/mujoco/solver_mujoco.py (1)
3093-3103: Nit:hfield_nameduplicatesname.
hfield_name(line 3096) is computed with the same format string asname(line 3038). You can just reusenamehere.Suggested simplification
- hfield_name = f"{model.shape_key[shape]}_{shape}" - spec.add_hfield( - name=hfield_name, + spec.add_hfield( + name=name, nrow=nrow, ncol=ncol, size=size, userdata=elevation_data, ) # Set geom to reference the hfield - geom_params["hfieldname"] = hfield_name + geom_params["hfieldname"] = namenewton/_src/utils/import_mjcf.py (1)
644-669:parse_meshesflag is overloaded to also gate heightfield parsing.Heightfield shapes are guarded by
parse_meshes, which semantically only applies to mesh geometry. A user callingparse_mjcf(..., parse_meshes=False)might not expect heightfields to also be skipped. This is a minor discoverability concern — consider renaming or adding a dedicatedparse_heightfieldsflag in a follow-up.The scale application on line 656 correctly addresses the earlier concern about unscaled heightfield sizes.
newton/_src/utils/heightfield.py (1)
82-82: Remove unusednoqadirective.Static analysis (Ruff RUF100) flags the
# noqa: PLC0415as unused sincePLC0415is not enabled. The lazy import of PIL is fine and appropriate here; the directive itself is just unnecessary.- from PIL import Image # noqa: PLC0415 + from PIL import Imagenewton/_src/geometry/types.py (1)
476-476: Remove unusednoqadirective.Ruff (RUF100) flags
# noqa: PLC0415as unused since that lint rule is not enabled.- from .inertia import compute_box_inertia # noqa: PLC0415 + from .inertia import compute_box_inertia
- Fix inertia computation: multiply half-extents by 2 for compute_box_inertia - Fix bounding sphere Z extent: use (size_z + size_base) / 2 - Use zero matrix instead of identity for static heightfield inertia - Use wp.config.verbose instead of undefined verbose variable - Add binary file header/payload validation in load_heightfield_elevation - Add inline elevation attribute parsing for MJCF hfield assets - Validate and pad size tuple to 4 elements in MJCF parsing - Apply global scale to heightfield size - Add type hints for Heightfield in utils.py and inertia.py - Simplify elevation loading: inline elevation > file > zeros - Add temp file cleanup in binary file test - Add test for inline elevation MJCF parsing Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Member
There are no newton heightfields yet, only a branch with experiments. I think it's fine to make the API work properly for Mujoco heightfields and then the Newton Heightfields adjust. |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@newton/_src/utils/import_mjcf.py`:
- Around line 332-342: The inline elevation parsing path currently stores raw
floats from hfield.attrib ("elevation") into elevation_arr and then
elevation_data without normalization; update the block that handles
elevation_str/elevation_arr in import_mjcf.py so that after creating
elevation_arr (and before reshaping/assigning elevation_data) you normalize
values to MuJoCo's [0,1] mapping using (arr - arr.min()) / (arr.max() -
arr.min() + small_eps) so edge cases avoid divide-by-zero; keep the size check
(arr.size == nrow * ncol) and then reshape into elevation_data as before; note
that PNG-loaded elevation already normalizes in load_heightfield_elevation so
only change the inline elevation branch.
🧹 Nitpick comments (5)
newton/_src/geometry/inertia.py (1)
536-558: Consider extracting the shared inertia-scaling logic into a helper.Lines 540–555 are an exact duplicate of the mesh branch at lines 562–578 (mass_ratio, component-wise tensor scaling, mat33 construction). A small helper such as
_scale_precomputed_inertia(src, scale, density)would eliminate this duplication and keep the two branches in sync if the formula ever changes.♻️ Sketch of a shared helper
+def _scale_precomputed_inertia( + src, scale: Vec3, density: float +) -> tuple[float, wp.vec3, wp.mat33]: + """Scale precomputed mass/com/inertia by (scale, density).""" + m, c, I = src.mass, src.com, src.I + s = wp.vec3(scale) + sx, sy, sz = s + mass_ratio = sx * sy * sz * density + m_new = m * mass_ratio + c_new = wp.cw_mul(c, s) + Ixx = I[0, 0] * (sy**2 + sz**2) / 2 * mass_ratio + Iyy = I[1, 1] * (sx**2 + sz**2) / 2 * mass_ratio + Izz = I[2, 2] * (sx**2 + sy**2) / 2 * mass_ratio + Ixy = I[0, 1] * sx * sy * mass_ratio + Ixz = I[0, 2] * sx * sz * mass_ratio + Iyz = I[1, 2] * sy * sz * mass_ratio + I_new = wp.mat33([[Ixx, Ixy, Ixz], [Ixy, Iyy, Iyz], [Ixz, Iyz, Izz]]) + return m_new, c_new, I_newThen both the HFIELD and MESH/SDF/CONVEX_MESH branches call
_scale_precomputed_inertia(src, scale, density).newton/_src/utils/import_mjcf.py (1)
668-702: Heightfield shape creation path looks correct.Elevation priority (inline → file → zeros), scale application, and
Heightfieldconstruction are all properly handled. Theadd_shape_heightfieldcall follows the same**shape_kwargspattern as other geometry types.One minor note: the
parse_meshesflag on line 668 also gates heightfield parsing, but the parameter's docstring (line 181) only mentions"mesh"type geometries. Consider updating the docstring to mention heightfield types as well.newton/tests/test_heightfield.py (3)
213-257: Clean upnoqadirectives and narrow the exception catch.Per Ruff, the
noqa: F401, PLC0415directive on line 216 is unnecessary (rules not enabled). Same applies to line 262.The broad
except Exceptionon line 256 can mask the specific failure reason;self.fail()will re-raise asAssertionErroranyway. Consider letting the exception propagate naturally — if the solver constructor raises, the test already fails with a useful traceback.♻️ Suggested simplification
- try: - import mujoco # noqa: F401, PLC0415 - except ImportError: - self.skipTest("MuJoCo not installed") + try: + import mujoco + except ImportError: + self.skipTest("MuJoCo not installed") ... - # Create MuJoCo solver (this will convert heightfield to MuJoCo format) - try: - newton.solvers.SolverMuJoCo(model) - except Exception as e: - self.fail(f"Failed to create MuJoCo solver with heightfield: {e}") + # Create MuJoCo solver (this will convert heightfield to MuJoCo format) + newton.solvers.SolverMuJoCo(model)
259-310: Collision assertion is very lenient — consider tightening.The check
final_z > -sphere_radiusonly verifies the sphere center didn't go below the negative of its own radius, which would mean the sphere fell completely through the heightfield. A stricter bound would verify it actually rests near the surface. For a flat heightfield withsize_base=0.01, the resting z should be approximately0.01 + sphere_radius(≈ 0.11).♻️ Suggested tighter bound
- self.assertGreater( - final_z, - -sphere_radius, - f"Sphere fell through heightfield: z={final_z:.4f}", - ) + # Sphere should rest near the surface (size_base=0.01 + sphere_radius) + self.assertGreater( + final_z, + 0.0, + f"Sphere fell through heightfield: z={final_z:.4f}", + )
343-363: Radius computation test looks correct; remove unused noqa.The expected formula matches the
compute_shape_radiusimplementation for HFIELD. The# noqa: PLC0415on line 345 is flagged by Ruff as unnecessary and can be removed.- from newton._src.geometry.utils import compute_shape_radius # noqa: PLC0415 + from newton._src.geometry.utils import compute_shape_radius
Contributor
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Fix all issues with AI agents
In `@newton/_src/geometry/inertia.py`:
- Around line 539-554: Rename the ambiguous variable I to a clearer name like
inertia across the block that handles precomputed inertia (the branch checking
src is not None and src.has_inertia and src.mass > 0.0); replace uses m, c, I =
src.mass, src.com, src.I with m, c, inertia = src.mass, src.com, src.I and then
update all subsequent references (Ixx, Iyy, Izz, Ixy, Ixz, Iyz calculations and
the construction of I_new) to index into inertia instead of I so the computed
I_new and other variables remain correct and ruff E741 is resolved.
In `@newton/_src/utils/heightfield.py`:
- Around line 23-27: Rename the internal helper function
load_heightfield_elevation to _load_heightfield_elevation to signal non-public
API; update its definition in heightfield.py and update all call sites (e.g.,
import_mjcf.py) to use _load_heightfield_elevation, and ensure any imports or
re-exports that referenced load_heightfield_elevation are adjusted or removed so
the public API no longer exposes it.
🧹 Nitpick comments (1)
newton/_src/utils/heightfield.py (1)
47-47: Remove unusednoqadirective.Ruff reports
noqa: PLC0415is not suppressing any enabled rule. Drop it to keep lint output clean.- from PIL import Image # noqa: PLC0415 + from PIL import Image
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@newton/_src/sim/builder.py`:
- Around line 4569-4615: The add_shape_heightfield method accepts
heightfield=None and passes it to add_shape, which can cause opaque failures;
add an explicit validation at the start of add_shape_heightfield (before using
self.default_shape_cfg or calling add_shape) that checks if heightfield is None
(and optionally validate its type/class if a Heightfield class exists) and
raises a clear ValueError like "heightfield must be provided" so callers get an
immediate, descriptive error instead of downstream exceptions from
add_shape/radius/inertia helpers.
🧹 Nitpick comments (1)
newton/_src/geometry/types.py (1)
439-491: Inconsistent default inertia betweenHeightfieldandMesh.When
compute_inertia=False,Mesh.__init__setsself.I = wp.mat33(np.eye(3))andself.mass = 1.0, whileHeightfieldsetsself.I = wp.mat33()(zero matrix) andself.mass = 0.0. The difference inmassis intentional (heightfields are static), but a zero inertia matrix could cause division-by-zero if any downstream code inverts it. Worth a brief comment explaining the choice, or using identity likeMeshdoes.Also, line 476: the
# noqa: PLC0415directive is flagged as unused by the linter and can be removed.Suggested fix for the noqa directive
- from .inertia import compute_box_inertia # noqa: PLC0415 + from .inertia import compute_box_inertia
adenzler-nvidia
left a comment
Member
There was a problem hiding this comment.
some comments. We should make sure the API is good enough.
@nvtw how would you judge effort of adding heightfield support to Newton? I'm a bit hesitant to add newton API for something MuJoCo-specific.
Member
For Newton support, I would just add one method to the builder which converts the height field to a triangle mesh (under the hood, the user does not need to know, we can optimize later), that's easy and fast enough. We can add full height field support later. |
Member
Author
|
@nvtw |
Member
Yes. Whenever the mujoco native collisions are not used and the contacts come from the Newton CollisionPipeline, the heightfield could just be converted to a triangle mesh and then one should still get proper collisions. We can do heightfield specific optimizations later. |
Remove compute_inertia and is_solid parameters from Heightfield. Heightfields are always static terrain with zero mass and inertia. Remove body parameter from add_shape_heightfield (always world-static). Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Member
Author
@nvtw I'm not sure how the builder can know where the contacts are generated. Can you elaborate? |
Member
Author
|
How about we add |
Treat heightfields as procedural meshes: grid cells produce triangles on demand via O(1) lookup (no BVH needed). Reuses existing GJK/MPR infrastructure for triangle-vs-convex contact generation. - Add HeightfieldData struct and get_triangle_from_heightfield_cell() - Populate shape_heightfield_data and elevation array in builder - Add HFIELD to broad phase AABB bounding sphere fallback - Route heightfield pairs in narrow phase primitive kernel - Skip unsupported pairs (hfield-hfield, hfield-plane, hfield-mesh/SDF) - Add heightfield midphase kernel (bounding sphere grid cell overlap) - Add heightfield triangle contacts kernel (GJK/MPR per cell triangle) - Refactor NarrowPhase counter array for cleaner feature-based layout - Add native collision pipeline tests Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
- Add HFIELD rendering to viewer by converting to mesh via create_mesh_heightfield (with row/col transpose for ij indexing) - Add basic_heightfield example with wave terrain and dropping spheres - Example supports --solver xpbd (native collision) and --solver mujoco Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
nvtw
reviewed
Feb 13, 2026
nvtw
reviewed
Feb 13, 2026
nvtw
previously approved these changes
Feb 13, 2026
Member
|
let's have @eric-heiden and @mmacklin approve the API additions as well. |
Member
|
In theory the heightfield could internally be stored as a mesh where the vertices are laid out row by row to allow 2d indexing and the z coordinate per vertex would be the heightfield height. This would give us free rendering and we might need less arrays in kernels. But I'm not sure if it is a good idea, the implementation as it is is fine too. |
adenzler-nvidia
left a comment
Member
There was a problem hiding this comment.
ok I think the API changes are actually so straightforward that we don't need additional approvals. Please address the comments and coderabbit, then let's ship this!
- Fix copyright year to 2026 in heightfield.py and test_heightfield.py - Remove unit reference (meters) from Heightfield docstring - Clarify hx/hy are already half-extents in bounding sphere comment - Add None validation for heightfield in add_shape_heightfield() Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
nvtw
approved these changes
Feb 16, 2026
This was referenced Feb 24, 2026
vastsoun
added a commit
to vastsoun/newton
that referenced
this pull request
Mar 3, 2026
* [Warp Raytrace] Added device parameter (newton-physics#1544) * [Warp Raytrace] Added device parameter to previously overlooked call (newton-physics#1545) * SolverMuJoCo: Fix tolerance clamping in update_solver_options_kernel (newton-physics#1546) * Change default shape_ke to align with MuJoCo, parse geom solref from MJCF for contact stiffness/damping (newton-physics#1491) Signed-off-by: Alain Denzler <adenzler@nvidia.com> * Fix log_shapes broadcasting for length-1 warp arrays (newton-physics#1550) * Fix XPBD restitution particle index (newton-physics#1557) * Out-of-Bound memory read in example_diffsim_bear newton-physics#1386 (newton-physics#1533) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add versioned documentation deployment to GitHub Pages (newton-physics#1560) * Fix broken documentation links after versioned docs deployment (newton-physics#1566) * VBD New Features (newton-physics#1479) Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com> * Add banners to membership verification workflow steps (newton-physics#1569) * Support cable junctions (newton-physics#1519) Signed-off-by: JC <jumyungc@nvidia.com> * Rename parameter I to inertia newton-physics#1543 (newton-physics#1551) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix example_robot_anymal_c_walk.py (newton-physics#1574) * Change everywhere linesearch to iterative (newton-physics#1573) * Remove standard collision pipeline (newton-physics#1538) * USD Plumbing MJC solver attributes through resolver and custom attribute framework (newton-physics#1463) Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> Signed-off-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Fix child shape filtering (newton-physics#1559) * Fix ViewerRerun ignoring hidden parameter in log_mesh and log_instances (newton-physics#1555) * Make NxN and SAP broad phase respect filtered pairs (newton-physics#1554) * Add --quiet flag to examples to suppress Warp messages (newton-physics#1585) * Defer resolution of MESH_MAXHULLVERT default in importers (newton-physics#1587) * Fix TypeError when finalizing SDF geometry with device kwarg (newton-physics#1586) * Make MESH_MAXHULLVERT a static class attribute Mesh.MAX_HULL_VERTICES. (newton-physics#1598) * Significant non-determinism in unified collision pipeline for anymal_c_walking example newton-physics#1505 (newton-physics#1588) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add test for non-contiguous case (newton-physics#1549) * Fix nightly Warp CI to resolve pre-release builds (newton-physics#1606) * Verify default class and value handling (newton-physics#1556) * SolverMuJoCo: Expand geom_margin to avoid OOB read with heterogeneous worlds (newton-physics#1607) * Fix bug in control clear method (newton-physics#1602) * Enable Use of Newton IK in Lab newton-physics#662 (newton-physics#1539) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix import of non-articulated joints (newton-physics#1535) * Deduplicate _process_joint_custom_attributes frequency handling (newton-physics#1584) * Add CI check for stale API docs and fix local build warnings (newton-physics#1570) * Update Pillow 12.0.0 to 12.1.1 (newton-physics#1612) * Prepare handling of mimic constraints in Newton (newton-physics#1523) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Support floating, base_joint and parent_body arguments for importers (newton-physics#1498) * Fix contact buffer memory overestimation (newton-physics#1614) * Configure banned-module-level-imports for ruff (newton-physics#1583) * Explicit `Contacts` instantiation with `Model.contacts()` and `CollisionPipeline.contacts()` (newton-physics#1445) * Fix the quadruped benchmark regression (newton-physics#1615) * Change default ignore_inertial_definitions from True to False (newton-physics#1537) * Finalize the Recording API (newton-physics#1600) * SolverMuJoCo: Fix ccd_iterations default (newton-physics#1631) * update gitignore to ignore Claude Code sandbox files (newton-physics#1628) * Add mimic joint support to SolverMuJoCo (newton-physics#1627) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add --no-cache-clear flag to test runner (newton-physics#1629) * Update MuJoCo and MuJoCo Warp to 3.5.0 release (newton-physics#1633) * Improve inertia parsing from USD (newton-physics#1605) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Remove standalone .typos.toml in favor of pyproject.toml config (newton-physics#1642) * Heightfield support newton-physics#1189 (newton-physics#1547) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Example_robot_policy: Replace ValueError with clean error for missing PhysX policy (newton-physics#1636) Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> * Avoid unnecessary inflation of the contact reduction voxel aabb (newton-physics#1650) * Rename num_worlds to world_count (newton-physics#1634) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Support parsing autolimits from MJCF (newton-physics#1651) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix particle-shape restitution ignoring body velocity (newton-physics#1273) (newton-physics#1580) * Add overflow warnings for narrow-phase collision buffers (newton-physics#1643) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Documentation: add units to model/state docstrings (newton-physics#1649) * fix(viewer): add missing JointType.BALL support to contact line kernel (newton-physics#1640) * Make terrain mesh visual-only in anymal C walking example (newton-physics#1660) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Fix initialization of collider state in MPM finite difference mode (newton-physics#1652) * docs: document ModelBuilder.default_shape_cfg (newton-physics#1662) * Finalize the collision API (newton-physics#1581) * Remove hardcoded subnet ID from AWS workflow (newton-physics#1664) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Attempt to fix AWS config (newton-physics#1666) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Update AWS workflows to g7e.2xlarge with multi-AZ failover (newton-physics#1669) * fix(viewer-usd): disambiguate log_points colors for N=3 warp arrays (newton-physics#1661) * Viewer gl optimizations (newton-physics#1656) Signed-off-by: Miles Macklin <mmacklin@nvidia.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: Eric Heiden <eheiden@nvidia.com> * docs: add articulation workflow guidance and regression check (newton-physics#1663) * fix(examples): propagate IK solution to model state in Franka example (newton-physics#1637) * fix(deps,docs): bump nbconvert to 7.17.0 and fix ArticulationView doctest (newton-physics#1670) * Cleanup and improve some example (newton-physics#1625) Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Handle zero-mass bodies and flip ensure_nonstatic_links default (newton-physics#1635) * Additional testing for ArticulationView (newton-physics#1527) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Update warp-lang nightly to 1.12.0.dev20260217 (newton-physics#1677) * Change default friction coefficients to match MuJoCo (newton-physics#1681) * Refactor mesh creation functions (newton-physics#1654) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Parse joint frictionloss from MJCF (newton-physics#1680) * Fix free joint body_pos and add ref/qpos0 support for MuJoCo solver (newton-physics#1645) * Fix root shapes in ArticulationView with fixed base (newton-physics#1639) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Use model collision methods and remove `create_collision_pipeline` from examples (newton-physics#1648) Co-authored-by: nvtw <110816143+nvtw@users.noreply.github.com> * Fix XPBD apply_joint_forces ignoring child joint transform (newton-physics#1582) * Adjust SDF API (newton-physics#1644) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Optimize test suite runtime (~18% faster) (newton-physics#1689) * Remove ensure_nonstatic_links option from importers (newton-physics#1682) Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> * SolverMuJoCo: Add geom_margin support, align thickness default with MuJoCo and schemas (newton-physics#1653) * refactor: privatize non-public solver internals (newton-physics#1683) * Fix option parsing with multiple <option> elements from includes (newton-physics#1692) * Bump warp-lang nightly and newton-usd-schemas (newton-physics#1693) * Get rid of tkinter dependency (newton-physics#1676) * Fix SDF example contact buffer overflow (newton-physics#1695) * Fix implicit biastype for position/velocity actuator shortcuts (newton-physics#1678) * Fix include processor to respect meshdir/texturedir (newton-physics#1685) * Reduce cold-cache Warp compile time for geometry modules (newton-physics#1618) Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Fix xyzw-to-wxyz quaternion conversion in body inertia kernel (newton-physics#1694) * Rename key to label and add hierarchical labels (newton-physics#1592) (newton-physics#1632) * Expose geometry SDF helpers on public API (newton-physics#1684) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Improve custom frequency handling from USD, parse MuJoCo actuators and tendons (newton-physics#1510) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Collapse fixed joints with non articulated bodies (newton-physics#1608) * Fix renaming joint_key -> joint_label (newton-physics#1700) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Bump .python-version from 3.11 to 3.12 (newton-physics#1702) * Replace CITATION.md with CITATION.cff (newton-physics#1706) * Respect MJCF contype=conaffinity=0 via collision_group=0 (newton-physics#1703) * Make ViewerUSD reuse existing USD layers for the same output path (newton-physics#1704) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Remove dead up-axis conversion from MuJoCo solver (newton-physics#1707) * Skip IK cube stacking example test (newton-physics#1713) * Fix global pairs (world=-1) not exported to MuJoCo spec (newton-physics#1705) * Reduce the memory consumption of hydroelastic contacts (newton-physics#1609) * Fix flakiness cube stacking (newton-physics#1714) * Fix collision shapes not toggleable in viewer UI (newton-physics#1715) * Fix softbody examples table layout in README (newton-physics#1716) * Standardize sensor APIs: label matching, keyword args, and update() method naming (newton-physics#1665) * Fix intermittent crash in parallel test runner from Manager proxy race (newton-physics#1721) * Clean up inertia.py function arguments to match Mesh.create_* API (newton-physics#1719) * Reduce default test runner verbosity (newton-physics#1723) * Add menagerie comparison tests for SolverMuJoCo (newton-physics#1720) Signed-off-by: Alain Denzler <adenzler@nvidia.com> Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> * Add spatial tendon support for MuJoCo solver (newton-physics#1687) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Expose qfrc_actuator from mujoco (newton-physics#1698) * Skip non-MODEL custom attributes in finalize validation (newton-physics#1734) * Resolve inheritrange for position actuators in MJCF parser (newton-physics#1727) * Support dampratio for position/velocity actuator shortcuts (newton-physics#1722) * Limit concurrency to 1 (newton-physics#1736) * Add a helper method for checking applied usd (newton-physics#1731) * Enhance playback URL handling in ViewerViser (newton-physics#1742) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Removed RenderShapeType (newton-physics#1748) * Override only MassAPI attributes that have been authored (newton-physics#1688) * Integration of newton-actuators (newton-physics#1342) Signed-off-by: jvonmuralt <jvonmuralt@nvidia.com> * Fix ArticulationView crash with fixed-joint-only articulations (newton-physics#1726) * Improve retrieval of Jupyter base URL in ViewerViser (newton-physics#1750) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Disable dynamics testing for UR5e and Apollo tests to avoid CI flakiness (newton-physics#1755) * Margin and Gap rename (newton-physics#1732) * Vbd Demos Fixing (newton-physics#1740) * Fix ViewerViser.log_lines method (newton-physics#1764) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Update warp-lang dependency to 1.12.0rc1 (newton-physics#1763) * Fix fromto capsule/cylinder orientation in MJCF parser (newton-physics#1741) * fix: multi-world particle BVH indexing (newton-physics#1641) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Clean up unused and internal-only kwargs in SolverMuJoCo (newton-physics#1766) * Parsing of the mimic joint and contact gap/margin from newton schemas (newton-physics#1690) Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> * API Refactor v2 (newton-physics#1749) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Support explicit geom mass attributes in MJCF (newton-physics#1744) * Bump flask and werkzeug in lockfile for security (newton-physics#1769) Co-authored-by: Cursor <cursoragent@cursor.com> * Split MJCF worldbody root bodies into separate articulations (newton-physics#1754) * Expose VBD rigid contact forces for solver coupling (newton-physics#1745) Signed-off-by: JC <jumyungc@nvidia.com> * Add MJCF ellipsoid geom import and regression test (newton-physics#1772) Co-authored-by: Cursor <cursoragent@cursor.com> * Weld equality constraints parsed from mjcf are given Nan as the default value of torquescale. The correct default should be 1.0 (newton-physics#1760) * Improve picking accuracy and stability (newton-physics#1712) * Franka cloth demo improvement (newton-physics#1765) * Support computing sensing object transforms & API cleanup (newton-physics#1759) * Remove threading workaround (newton-physics#1751) * [Warp Raytrace] Consolidated ray intersect functions, renamed Options to Config (newton-physics#1767) * Improve README example gallery for PyPI compatibility (newton-physics#1776) * Fix issue with mesh in rerun viewer (newton-physics#1768) * Add PhysxMimicJointAPI parsing to USD importer (newton-physics#1735) * Move some math functions to Warp (newton-physics#1717) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Add test to ensure MJCF xform argument is relative (newton-physics#1777) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Emit diaginertia instead of fullinertia for diagonal body inertia (newton-physics#1780) * Change default joint armature from 0.01 to 0 (newton-physics#1782) * Fix default kp/kv for position and velocity actuators (newton-physics#1786) * Lock body inertia after explicit MJCF <inertial> element (newton-physics#1784) * Fix for MJCF actuator custom attributes (newton-physics#1783) * Bump version to 0.2.3 Prepare the package metadata for the v0.2.3 release tag. * Fix ViewerRerun rendering of instances from hidden meshes (newton-physics#1788) * API cleanup (newton-physics#1789) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * BODY actuator target name bypasses body name de-duplication in SolverMuJoCo (newton-physics#1729) * Use default density for visual geoms in MJCF import (newton-physics#1781) * Fix GL viewer crash on Wayland (newton-physics#1793) * Make USD xform parameter control absolute articulation placement (newton-physics#1771) * Fix CUDA context corruption in SDF implementation (newton-physics#1792) * Bump mujoco-warp dependency to 3.5.0.2 (newton-physics#1779) * Fix MuJoCo margin/gap conversion (newton-physics#1785) * Bump version to 1.1.0.dev0 (newton-physics#1798) * Missing unittest.main added back to test_import_mjcf.py. Helps with F5 debugging in VS Code. (newton-physics#1796) * Improve H1 example (newton-physics#1801) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Fix ViewerViser.set_camera() (newton-physics#1805) * Rename examples to follow prefix-first naming convention (newton-physics#1802) * Improve MuJoCo actuator domain randomization (newton-physics#1773) * Restore in_cup test in hydro example (newton-physics#1775) * Fix `newton.geometry` imports and change of Mesh maxhullverts global constant * Adapt to having both margin and gap arrays for each geom/shape * Fix for newton.geometry API changes in primitive/narrowphase.py * Fix removal of `BroadPhaseMode` IntEnum and revert to newton-sytle string literals * WIP: Adapt geometry/unified.py to fix breaking changes to `newton.geometry` API * First pass of API adjustments * Patch more gaps (with respect to margin and gap renaming) * GeoType.SDF was removed - reflect that in Kamino * Introduce CoM position offsets w.r.t body frame and operations to convert between body CoM state and generic local body-fixed reference frames. * Add caching of per-entity labels/names/keys in the Model subcontainers * Remove SDF shape wrapper since it's now internal to mesh handling in CD pipelines * Fix USD test assets * Add Newton <--> Kamino joint type conversion operations and per-space default limit constants * Add some cleanup to geometry and unified CD + UTs * Add Newton <--> Kamino shape type conversion operations * Apply new default joint coord limit constants to limits.py * Adapt foubrar model builder and USD asset to produce the same result in sim example * Purge "physical" goems and collapse all into a single group, and purge geometry "layers" * Disable allocation per-joint wrenches by default and make them optional * Make `Model` a dataclass * Separate ModelData* containers into own `core/data.py` module * Fix imports of ModelData * Rename `Model` as `ModelKamino` * Rename `ModelData` as `DataKamino` * Rename `State` as `StateKamino` * Rename `Control` as `ControlKamino` * Rename `Limits` as `LimitsKamino` * Rename `Contacts` as `ContactsKamino` * Rename `ModelBuilder` as `ModelBuilderKamino` * Make imports in test utilities relative * Revise CD meta-data attributes and their computation in GeometryModel and ModelBuilderKamino * Revise primitive CD pipeline * Revise unified CD pipeline * Revise CD front-end interfaces * Fix UTs and relevant utils for interface changes to CD * Change to `wp.DeviceLike` to account for upcoming deprecation of `Devicelike` * Depracate legacy HDF5 data io (will be replaced in the future) * Fix banned imports at module level * Modify USD importer to detect articulations and order geoms and joints similarly to how the Newton `parse_usd` function does. * Add conversion operation from `newton.Model` to `ModelKamino` * Add data, state and control container conversions * Add SolverKamino wrapper that fullfils newton integration interface * Add newton integration examples * Add SolverKamino to newton solver module imports * WIP: fix problem with lambda_j being allocated for only kinematic constraints and failing on array copying * Fix banned git import in benchmark * Rename *Settings to *Config (#213) * Rename SolverKaminoSettings -> SolverKaminoConfig * Rename DualProblemSettings -> DualProblemConfig * Rename CollisionDetectorSettings -> CollisionDetectorConfig * Rename PADMMSettings -> PADMMConfig * Rename ForwardKinematicsSolverSettings -> ForwardKinematicsSolverConfig * Rename SimulatorSettings -> SimulatorConfig * Add check for model compatibility in SolverKamino (#209) * Fix device assignment in sparse CG test on CPU (#216) * Replace Enum-type config attributes with Literal (#215) * Replace warmstart mode config param with literal * Replace contact warmstart mode config param with literal * Replace rotation correction config param with literal * Replace penalty update config param with literal * Replace FK preconditioner option config param with literal * Add post-init checks for dual/PADMM configs * Rename FKPreconditionerOptions to FKPreconditionerType * Remove WorldDescriptor from ModelKamino (#219) * Add geom index offset array to model info * Replace access to world description in model * Remove world descriptor from model * Fix computation of kinematics residual with sparse Jacobian (#220) * Migrates `ModelKaminoSize` to it's own module to avoid circular dependency between core/model.py and core/state.py and rename it as `SizeKamino`. * WIP: Fix SolverKamino wrapper * Fix circular dependency in conversions.py * Fix some broken unit tests and WIP to fix fourbar contact conversions and rendering * Make some conversions @staticmethods instead, because they dont need common class attributes * Fix declaration of custom state attributes and their conversion to/from newton and kamino * WIP: Debug model conversion and newton sim examples * Rename and cleanup start index array of per-world geoms * Model conversion and newton sim examples now work. * Make gravity conversion operation re-usable * Migrates boxes_fourbar builder using newton.ModelBuilder to it's own module to prepare for migration. * Use gravity conversion utility func in SolverKamino * Add reusable joint-parameterization conversion utility * Remove world-descriptor from model converter * Rename helper converter that handles entity-local transforms * Add some cleanup to DR Legs, ANYmal D and basic four-bar examples * Fix module-level imports of additional kamino-specific development dependencies * Fix module-level imports of additional kamino-specific development dependencies * Fix erroneous merge conflict. --------- Signed-off-by: Alain Denzler <adenzler@nvidia.com> Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> Signed-off-by: JC <jumyungc@nvidia.com> Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> Signed-off-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Miles Macklin <mmacklin@nvidia.com> Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> Signed-off-by: jvonmuralt <jvonmuralt@nvidia.com> Co-authored-by: Daniela Hase <116915287+daniela-hase@users.noreply.github.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> Co-authored-by: Eric Shi <97630937+shi-eric@users.noreply.github.com> Co-authored-by: Anka Chen <AnkaChan@users.noreply.github.com> Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com> Co-authored-by: JC-nvidia <116605903+jumyungc@users.noreply.github.com> Co-authored-by: Kenny Vilella <kvilella@nvidia.com> Co-authored-by: nvtw <110816143+nvtw@users.noreply.github.com> Co-authored-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: Lennart Röstel <65088822+lenroe@users.noreply.github.com> Co-authored-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: jvonmuralt <jvonmuralt@nvidia.com> Co-authored-by: camevor <camevor@nvidia.com> Co-authored-by: mzamoramora-nvidia <mzamoramora@nvidia.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Alessandro Roncone <alecive87@gmail.com> Co-authored-by: gdaviet <57617656+gdaviet@users.noreply.github.com> Co-authored-by: Miles Macklin <mmacklin@nvidia.com> Co-authored-by: Gordon Yeoman <gyeomannvidia@users.noreply.github.com> Co-authored-by: Lukasz Wawrzyniak <lwawrzyniak@nvidia.com> Co-authored-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Lorenzo Terenzi <lorenzoterenzi96@gmail.com> Co-authored-by: smollerNV <164020096+smollerNV@users.noreply.github.com> Co-authored-by: twidmer <twidmer@nvidia.com> Co-authored-by: Christian Schumacher <christian.schumacher@disney.com> Co-authored-by: Guirec-Maloisel <25688871+Guirec-Maloisel@users.noreply.github.com>
vastsoun
added a commit
to vastsoun/newton
that referenced
this pull request
Mar 4, 2026
* [Warp Raytrace] Added device parameter (newton-physics#1544) * [Warp Raytrace] Added device parameter to previously overlooked call (newton-physics#1545) * SolverMuJoCo: Fix tolerance clamping in update_solver_options_kernel (newton-physics#1546) * Change default shape_ke to align with MuJoCo, parse geom solref from MJCF for contact stiffness/damping (newton-physics#1491) Signed-off-by: Alain Denzler <adenzler@nvidia.com> * Fix log_shapes broadcasting for length-1 warp arrays (newton-physics#1550) * Fix XPBD restitution particle index (newton-physics#1557) * Out-of-Bound memory read in example_diffsim_bear newton-physics#1386 (newton-physics#1533) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add versioned documentation deployment to GitHub Pages (newton-physics#1560) * Fix broken documentation links after versioned docs deployment (newton-physics#1566) * VBD New Features (newton-physics#1479) Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com> * Add banners to membership verification workflow steps (newton-physics#1569) * Support cable junctions (newton-physics#1519) Signed-off-by: JC <jumyungc@nvidia.com> * Rename parameter I to inertia newton-physics#1543 (newton-physics#1551) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix example_robot_anymal_c_walk.py (newton-physics#1574) * Change everywhere linesearch to iterative (newton-physics#1573) * Remove standard collision pipeline (newton-physics#1538) * USD Plumbing MJC solver attributes through resolver and custom attribute framework (newton-physics#1463) Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> Signed-off-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Fix child shape filtering (newton-physics#1559) * Fix ViewerRerun ignoring hidden parameter in log_mesh and log_instances (newton-physics#1555) * Make NxN and SAP broad phase respect filtered pairs (newton-physics#1554) * Add --quiet flag to examples to suppress Warp messages (newton-physics#1585) * Defer resolution of MESH_MAXHULLVERT default in importers (newton-physics#1587) * Fix TypeError when finalizing SDF geometry with device kwarg (newton-physics#1586) * Make MESH_MAXHULLVERT a static class attribute Mesh.MAX_HULL_VERTICES. (newton-physics#1598) * Significant non-determinism in unified collision pipeline for anymal_c_walking example newton-physics#1505 (newton-physics#1588) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add test for non-contiguous case (newton-physics#1549) * Fix nightly Warp CI to resolve pre-release builds (newton-physics#1606) * Verify default class and value handling (newton-physics#1556) * SolverMuJoCo: Expand geom_margin to avoid OOB read with heterogeneous worlds (newton-physics#1607) * Fix bug in control clear method (newton-physics#1602) * Enable Use of Newton IK in Lab newton-physics#662 (newton-physics#1539) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix import of non-articulated joints (newton-physics#1535) * Deduplicate _process_joint_custom_attributes frequency handling (newton-physics#1584) * Add CI check for stale API docs and fix local build warnings (newton-physics#1570) * Update Pillow 12.0.0 to 12.1.1 (newton-physics#1612) * Prepare handling of mimic constraints in Newton (newton-physics#1523) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Support floating, base_joint and parent_body arguments for importers (newton-physics#1498) * Fix contact buffer memory overestimation (newton-physics#1614) * Configure banned-module-level-imports for ruff (newton-physics#1583) * Explicit `Contacts` instantiation with `Model.contacts()` and `CollisionPipeline.contacts()` (newton-physics#1445) * Fix the quadruped benchmark regression (newton-physics#1615) * Change default ignore_inertial_definitions from True to False (newton-physics#1537) * Finalize the Recording API (newton-physics#1600) * SolverMuJoCo: Fix ccd_iterations default (newton-physics#1631) * update gitignore to ignore Claude Code sandbox files (newton-physics#1628) * Add mimic joint support to SolverMuJoCo (newton-physics#1627) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add --no-cache-clear flag to test runner (newton-physics#1629) * Update MuJoCo and MuJoCo Warp to 3.5.0 release (newton-physics#1633) * Improve inertia parsing from USD (newton-physics#1605) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Remove standalone .typos.toml in favor of pyproject.toml config (newton-physics#1642) * Heightfield support newton-physics#1189 (newton-physics#1547) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Example_robot_policy: Replace ValueError with clean error for missing PhysX policy (newton-physics#1636) Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> * Avoid unnecessary inflation of the contact reduction voxel aabb (newton-physics#1650) * Rename num_worlds to world_count (newton-physics#1634) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Support parsing autolimits from MJCF (newton-physics#1651) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix particle-shape restitution ignoring body velocity (newton-physics#1273) (newton-physics#1580) * Add overflow warnings for narrow-phase collision buffers (newton-physics#1643) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Documentation: add units to model/state docstrings (newton-physics#1649) * fix(viewer): add missing JointType.BALL support to contact line kernel (newton-physics#1640) * Make terrain mesh visual-only in anymal C walking example (newton-physics#1660) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Fix initialization of collider state in MPM finite difference mode (newton-physics#1652) * docs: document ModelBuilder.default_shape_cfg (newton-physics#1662) * Finalize the collision API (newton-physics#1581) * Remove hardcoded subnet ID from AWS workflow (newton-physics#1664) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Attempt to fix AWS config (newton-physics#1666) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Update AWS workflows to g7e.2xlarge with multi-AZ failover (newton-physics#1669) * fix(viewer-usd): disambiguate log_points colors for N=3 warp arrays (newton-physics#1661) * Viewer gl optimizations (newton-physics#1656) Signed-off-by: Miles Macklin <mmacklin@nvidia.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: Eric Heiden <eheiden@nvidia.com> * docs: add articulation workflow guidance and regression check (newton-physics#1663) * fix(examples): propagate IK solution to model state in Franka example (newton-physics#1637) * fix(deps,docs): bump nbconvert to 7.17.0 and fix ArticulationView doctest (newton-physics#1670) * Cleanup and improve some example (newton-physics#1625) Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Handle zero-mass bodies and flip ensure_nonstatic_links default (newton-physics#1635) * Additional testing for ArticulationView (newton-physics#1527) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Update warp-lang nightly to 1.12.0.dev20260217 (newton-physics#1677) * Change default friction coefficients to match MuJoCo (newton-physics#1681) * Refactor mesh creation functions (newton-physics#1654) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Parse joint frictionloss from MJCF (newton-physics#1680) * Fix free joint body_pos and add ref/qpos0 support for MuJoCo solver (newton-physics#1645) * Fix root shapes in ArticulationView with fixed base (newton-physics#1639) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Use model collision methods and remove `create_collision_pipeline` from examples (newton-physics#1648) Co-authored-by: nvtw <110816143+nvtw@users.noreply.github.com> * Fix XPBD apply_joint_forces ignoring child joint transform (newton-physics#1582) * Adjust SDF API (newton-physics#1644) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Optimize test suite runtime (~18% faster) (newton-physics#1689) * Remove ensure_nonstatic_links option from importers (newton-physics#1682) Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> * SolverMuJoCo: Add geom_margin support, align thickness default with MuJoCo and schemas (newton-physics#1653) * refactor: privatize non-public solver internals (newton-physics#1683) * Fix option parsing with multiple <option> elements from includes (newton-physics#1692) * Bump warp-lang nightly and newton-usd-schemas (newton-physics#1693) * Get rid of tkinter dependency (newton-physics#1676) * Fix SDF example contact buffer overflow (newton-physics#1695) * Fix implicit biastype for position/velocity actuator shortcuts (newton-physics#1678) * Fix include processor to respect meshdir/texturedir (newton-physics#1685) * Reduce cold-cache Warp compile time for geometry modules (newton-physics#1618) Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Fix xyzw-to-wxyz quaternion conversion in body inertia kernel (newton-physics#1694) * Rename key to label and add hierarchical labels (newton-physics#1592) (newton-physics#1632) * Expose geometry SDF helpers on public API (newton-physics#1684) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Improve custom frequency handling from USD, parse MuJoCo actuators and tendons (newton-physics#1510) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Collapse fixed joints with non articulated bodies (newton-physics#1608) * Fix renaming joint_key -> joint_label (newton-physics#1700) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Bump .python-version from 3.11 to 3.12 (newton-physics#1702) * Replace CITATION.md with CITATION.cff (newton-physics#1706) * Respect MJCF contype=conaffinity=0 via collision_group=0 (newton-physics#1703) * Make ViewerUSD reuse existing USD layers for the same output path (newton-physics#1704) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Remove dead up-axis conversion from MuJoCo solver (newton-physics#1707) * Skip IK cube stacking example test (newton-physics#1713) * Fix global pairs (world=-1) not exported to MuJoCo spec (newton-physics#1705) * Reduce the memory consumption of hydroelastic contacts (newton-physics#1609) * Fix flakiness cube stacking (newton-physics#1714) * Fix collision shapes not toggleable in viewer UI (newton-physics#1715) * Fix softbody examples table layout in README (newton-physics#1716) * Standardize sensor APIs: label matching, keyword args, and update() method naming (newton-physics#1665) * Fix intermittent crash in parallel test runner from Manager proxy race (newton-physics#1721) * Clean up inertia.py function arguments to match Mesh.create_* API (newton-physics#1719) * Reduce default test runner verbosity (newton-physics#1723) * Add menagerie comparison tests for SolverMuJoCo (newton-physics#1720) Signed-off-by: Alain Denzler <adenzler@nvidia.com> Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> * Add spatial tendon support for MuJoCo solver (newton-physics#1687) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Expose qfrc_actuator from mujoco (newton-physics#1698) * Skip non-MODEL custom attributes in finalize validation (newton-physics#1734) * Resolve inheritrange for position actuators in MJCF parser (newton-physics#1727) * Support dampratio for position/velocity actuator shortcuts (newton-physics#1722) * Limit concurrency to 1 (newton-physics#1736) * Add a helper method for checking applied usd (newton-physics#1731) * Enhance playback URL handling in ViewerViser (newton-physics#1742) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Removed RenderShapeType (newton-physics#1748) * Override only MassAPI attributes that have been authored (newton-physics#1688) * Integration of newton-actuators (newton-physics#1342) Signed-off-by: jvonmuralt <jvonmuralt@nvidia.com> * Fix ArticulationView crash with fixed-joint-only articulations (newton-physics#1726) * Improve retrieval of Jupyter base URL in ViewerViser (newton-physics#1750) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Disable dynamics testing for UR5e and Apollo tests to avoid CI flakiness (newton-physics#1755) * Margin and Gap rename (newton-physics#1732) * Vbd Demos Fixing (newton-physics#1740) * Fix ViewerViser.log_lines method (newton-physics#1764) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Update warp-lang dependency to 1.12.0rc1 (newton-physics#1763) * Fix fromto capsule/cylinder orientation in MJCF parser (newton-physics#1741) * fix: multi-world particle BVH indexing (newton-physics#1641) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Clean up unused and internal-only kwargs in SolverMuJoCo (newton-physics#1766) * Parsing of the mimic joint and contact gap/margin from newton schemas (newton-physics#1690) Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> * API Refactor v2 (newton-physics#1749) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Support explicit geom mass attributes in MJCF (newton-physics#1744) * Bump flask and werkzeug in lockfile for security (newton-physics#1769) Co-authored-by: Cursor <cursoragent@cursor.com> * Split MJCF worldbody root bodies into separate articulations (newton-physics#1754) * Expose VBD rigid contact forces for solver coupling (newton-physics#1745) Signed-off-by: JC <jumyungc@nvidia.com> * Add MJCF ellipsoid geom import and regression test (newton-physics#1772) Co-authored-by: Cursor <cursoragent@cursor.com> * Weld equality constraints parsed from mjcf are given Nan as the default value of torquescale. The correct default should be 1.0 (newton-physics#1760) * Improve picking accuracy and stability (newton-physics#1712) * Franka cloth demo improvement (newton-physics#1765) * Support computing sensing object transforms & API cleanup (newton-physics#1759) * Remove threading workaround (newton-physics#1751) * [Warp Raytrace] Consolidated ray intersect functions, renamed Options to Config (newton-physics#1767) * Improve README example gallery for PyPI compatibility (newton-physics#1776) * Fix issue with mesh in rerun viewer (newton-physics#1768) * Add PhysxMimicJointAPI parsing to USD importer (newton-physics#1735) * Move some math functions to Warp (newton-physics#1717) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Add test to ensure MJCF xform argument is relative (newton-physics#1777) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Emit diaginertia instead of fullinertia for diagonal body inertia (newton-physics#1780) * Change default joint armature from 0.01 to 0 (newton-physics#1782) * Fix default kp/kv for position and velocity actuators (newton-physics#1786) * Lock body inertia after explicit MJCF <inertial> element (newton-physics#1784) * Fix for MJCF actuator custom attributes (newton-physics#1783) * Bump version to 0.2.3 Prepare the package metadata for the v0.2.3 release tag. * Fix ViewerRerun rendering of instances from hidden meshes (newton-physics#1788) * API cleanup (newton-physics#1789) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * BODY actuator target name bypasses body name de-duplication in SolverMuJoCo (newton-physics#1729) * Use default density for visual geoms in MJCF import (newton-physics#1781) * Fix GL viewer crash on Wayland (newton-physics#1793) * Make USD xform parameter control absolute articulation placement (newton-physics#1771) * Fix CUDA context corruption in SDF implementation (newton-physics#1792) * Bump mujoco-warp dependency to 3.5.0.2 (newton-physics#1779) * Fix MuJoCo margin/gap conversion (newton-physics#1785) * Bump version to 1.1.0.dev0 (newton-physics#1798) * Missing unittest.main added back to test_import_mjcf.py. Helps with F5 debugging in VS Code. (newton-physics#1796) * Improve H1 example (newton-physics#1801) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Fix ViewerViser.set_camera() (newton-physics#1805) * Rename examples to follow prefix-first naming convention (newton-physics#1802) * Improve MuJoCo actuator domain randomization (newton-physics#1773) * Restore in_cup test in hydro example (newton-physics#1775) * Fix `newton.geometry` imports and change of Mesh maxhullverts global constant * Adapt to having both margin and gap arrays for each geom/shape * Fix for newton.geometry API changes in primitive/narrowphase.py * Fix removal of `BroadPhaseMode` IntEnum and revert to newton-sytle string literals * WIP: Adapt geometry/unified.py to fix breaking changes to `newton.geometry` API * First pass of API adjustments * Patch more gaps (with respect to margin and gap renaming) * GeoType.SDF was removed - reflect that in Kamino * Introduce CoM position offsets w.r.t body frame and operations to convert between body CoM state and generic local body-fixed reference frames. * Add caching of per-entity labels/names/keys in the Model subcontainers * Remove SDF shape wrapper since it's now internal to mesh handling in CD pipelines * Fix USD test assets * Add Newton <--> Kamino joint type conversion operations and per-space default limit constants * Add some cleanup to geometry and unified CD + UTs * Add Newton <--> Kamino shape type conversion operations * Apply new default joint coord limit constants to limits.py * Adapt foubrar model builder and USD asset to produce the same result in sim example * Purge "physical" goems and collapse all into a single group, and purge geometry "layers" * Disable allocation per-joint wrenches by default and make them optional * Make `Model` a dataclass * Separate ModelData* containers into own `core/data.py` module * Fix imports of ModelData * Rename `Model` as `ModelKamino` * Rename `ModelData` as `DataKamino` * Rename `State` as `StateKamino` * Rename `Control` as `ControlKamino` * Rename `Limits` as `LimitsKamino` * Rename `Contacts` as `ContactsKamino` * Rename `ModelBuilder` as `ModelBuilderKamino` * Make imports in test utilities relative * Revise CD meta-data attributes and their computation in GeometryModel and ModelBuilderKamino * Revise primitive CD pipeline * Revise unified CD pipeline * Revise CD front-end interfaces * Fix UTs and relevant utils for interface changes to CD * Change to `wp.DeviceLike` to account for upcoming deprecation of `Devicelike` * Depracate legacy HDF5 data io (will be replaced in the future) * Fix banned imports at module level * Modify USD importer to detect articulations and order geoms and joints similarly to how the Newton `parse_usd` function does. * Add conversion operation from `newton.Model` to `ModelKamino` * Add data, state and control container conversions * Add SolverKamino wrapper that fullfils newton integration interface * Add newton integration examples * Add SolverKamino to newton solver module imports * WIP: fix problem with lambda_j being allocated for only kinematic constraints and failing on array copying * Fix banned git import in benchmark * Rename *Settings to *Config (#213) * Rename SolverKaminoSettings -> SolverKaminoConfig * Rename DualProblemSettings -> DualProblemConfig * Rename CollisionDetectorSettings -> CollisionDetectorConfig * Rename PADMMSettings -> PADMMConfig * Rename ForwardKinematicsSolverSettings -> ForwardKinematicsSolverConfig * Rename SimulatorSettings -> SimulatorConfig * Add check for model compatibility in SolverKamino (#209) * Fix device assignment in sparse CG test on CPU (#216) * Replace Enum-type config attributes with Literal (#215) * Replace warmstart mode config param with literal * Replace contact warmstart mode config param with literal * Replace rotation correction config param with literal * Replace penalty update config param with literal * Replace FK preconditioner option config param with literal * Add post-init checks for dual/PADMM configs * Rename FKPreconditionerOptions to FKPreconditionerType * Remove WorldDescriptor from ModelKamino (#219) * Add geom index offset array to model info * Replace access to world description in model * Remove world descriptor from model * Fix computation of kinematics residual with sparse Jacobian (#220) * Migrates `ModelKaminoSize` to it's own module to avoid circular dependency between core/model.py and core/state.py and rename it as `SizeKamino`. * WIP: Fix SolverKamino wrapper * Fix circular dependency in conversions.py * Fix some broken unit tests and WIP to fix fourbar contact conversions and rendering * Make some conversions @staticmethods instead, because they dont need common class attributes * Fix declaration of custom state attributes and their conversion to/from newton and kamino * WIP: Debug model conversion and newton sim examples * Rename and cleanup start index array of per-world geoms * Model conversion and newton sim examples now work. * Make gravity conversion operation re-usable * Migrates boxes_fourbar builder using newton.ModelBuilder to it's own module to prepare for migration. * Use gravity conversion utility func in SolverKamino * Add reusable joint-parameterization conversion utility * Remove world-descriptor from model converter * Rename helper converter that handles entity-local transforms * Add some cleanup to DR Legs, ANYmal D and basic four-bar examples * Fix module-level imports of additional kamino-specific development dependencies * Fix module-level imports of additional kamino-specific development dependencies * Fix erroneous merge conflict. --------- Signed-off-by: Alain Denzler <adenzler@nvidia.com> Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> Signed-off-by: JC <jumyungc@nvidia.com> Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> Signed-off-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Miles Macklin <mmacklin@nvidia.com> Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> Signed-off-by: jvonmuralt <jvonmuralt@nvidia.com> Co-authored-by: Daniela Hase <116915287+daniela-hase@users.noreply.github.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> Co-authored-by: Eric Shi <97630937+shi-eric@users.noreply.github.com> Co-authored-by: Anka Chen <AnkaChan@users.noreply.github.com> Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com> Co-authored-by: JC-nvidia <116605903+jumyungc@users.noreply.github.com> Co-authored-by: Kenny Vilella <kvilella@nvidia.com> Co-authored-by: nvtw <110816143+nvtw@users.noreply.github.com> Co-authored-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: Lennart Röstel <65088822+lenroe@users.noreply.github.com> Co-authored-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: jvonmuralt <jvonmuralt@nvidia.com> Co-authored-by: camevor <camevor@nvidia.com> Co-authored-by: mzamoramora-nvidia <mzamoramora@nvidia.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Alessandro Roncone <alecive87@gmail.com> Co-authored-by: gdaviet <57617656+gdaviet@users.noreply.github.com> Co-authored-by: Miles Macklin <mmacklin@nvidia.com> Co-authored-by: Gordon Yeoman <gyeomannvidia@users.noreply.github.com> Co-authored-by: Lukasz Wawrzyniak <lwawrzyniak@nvidia.com> Co-authored-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Lorenzo Terenzi <lorenzoterenzi96@gmail.com> Co-authored-by: smollerNV <164020096+smollerNV@users.noreply.github.com> Co-authored-by: twidmer <twidmer@nvidia.com> Co-authored-by: Christian Schumacher <christian.schumacher@disney.com> Co-authored-by: Guirec-Maloisel <25688871+Guirec-Maloisel@users.noreply.github.com>
vastsoun
added a commit
to vastsoun/newton
that referenced
this pull request
Mar 4, 2026
* [Warp Raytrace] Added device parameter (newton-physics#1544) * [Warp Raytrace] Added device parameter to previously overlooked call (newton-physics#1545) * SolverMuJoCo: Fix tolerance clamping in update_solver_options_kernel (newton-physics#1546) * Change default shape_ke to align with MuJoCo, parse geom solref from MJCF for contact stiffness/damping (newton-physics#1491) Signed-off-by: Alain Denzler <adenzler@nvidia.com> * Fix log_shapes broadcasting for length-1 warp arrays (newton-physics#1550) * Fix XPBD restitution particle index (newton-physics#1557) * Out-of-Bound memory read in example_diffsim_bear newton-physics#1386 (newton-physics#1533) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add versioned documentation deployment to GitHub Pages (newton-physics#1560) * Fix broken documentation links after versioned docs deployment (newton-physics#1566) * VBD New Features (newton-physics#1479) Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com> * Add banners to membership verification workflow steps (newton-physics#1569) * Support cable junctions (newton-physics#1519) Signed-off-by: JC <jumyungc@nvidia.com> * Rename parameter I to inertia newton-physics#1543 (newton-physics#1551) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix example_robot_anymal_c_walk.py (newton-physics#1574) * Change everywhere linesearch to iterative (newton-physics#1573) * Remove standard collision pipeline (newton-physics#1538) * USD Plumbing MJC solver attributes through resolver and custom attribute framework (newton-physics#1463) Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> Signed-off-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Fix child shape filtering (newton-physics#1559) * Fix ViewerRerun ignoring hidden parameter in log_mesh and log_instances (newton-physics#1555) * Make NxN and SAP broad phase respect filtered pairs (newton-physics#1554) * Add --quiet flag to examples to suppress Warp messages (newton-physics#1585) * Defer resolution of MESH_MAXHULLVERT default in importers (newton-physics#1587) * Fix TypeError when finalizing SDF geometry with device kwarg (newton-physics#1586) * Make MESH_MAXHULLVERT a static class attribute Mesh.MAX_HULL_VERTICES. (newton-physics#1598) * Significant non-determinism in unified collision pipeline for anymal_c_walking example newton-physics#1505 (newton-physics#1588) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add test for non-contiguous case (newton-physics#1549) * Fix nightly Warp CI to resolve pre-release builds (newton-physics#1606) * Verify default class and value handling (newton-physics#1556) * SolverMuJoCo: Expand geom_margin to avoid OOB read with heterogeneous worlds (newton-physics#1607) * Fix bug in control clear method (newton-physics#1602) * Enable Use of Newton IK in Lab newton-physics#662 (newton-physics#1539) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix import of non-articulated joints (newton-physics#1535) * Deduplicate _process_joint_custom_attributes frequency handling (newton-physics#1584) * Add CI check for stale API docs and fix local build warnings (newton-physics#1570) * Update Pillow 12.0.0 to 12.1.1 (newton-physics#1612) * Prepare handling of mimic constraints in Newton (newton-physics#1523) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Support floating, base_joint and parent_body arguments for importers (newton-physics#1498) * Fix contact buffer memory overestimation (newton-physics#1614) * Configure banned-module-level-imports for ruff (newton-physics#1583) * Explicit `Contacts` instantiation with `Model.contacts()` and `CollisionPipeline.contacts()` (newton-physics#1445) * Fix the quadruped benchmark regression (newton-physics#1615) * Change default ignore_inertial_definitions from True to False (newton-physics#1537) * Finalize the Recording API (newton-physics#1600) * SolverMuJoCo: Fix ccd_iterations default (newton-physics#1631) * update gitignore to ignore Claude Code sandbox files (newton-physics#1628) * Add mimic joint support to SolverMuJoCo (newton-physics#1627) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Add --no-cache-clear flag to test runner (newton-physics#1629) * Update MuJoCo and MuJoCo Warp to 3.5.0 release (newton-physics#1633) * Improve inertia parsing from USD (newton-physics#1605) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Remove standalone .typos.toml in favor of pyproject.toml config (newton-physics#1642) * Heightfield support newton-physics#1189 (newton-physics#1547) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Example_robot_policy: Replace ValueError with clean error for missing PhysX policy (newton-physics#1636) Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> * Avoid unnecessary inflation of the contact reduction voxel aabb (newton-physics#1650) * Rename num_worlds to world_count (newton-physics#1634) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Support parsing autolimits from MJCF (newton-physics#1651) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Fix particle-shape restitution ignoring body velocity (newton-physics#1273) (newton-physics#1580) * Add overflow warnings for narrow-phase collision buffers (newton-physics#1643) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Documentation: add units to model/state docstrings (newton-physics#1649) * fix(viewer): add missing JointType.BALL support to contact line kernel (newton-physics#1640) * Make terrain mesh visual-only in anymal C walking example (newton-physics#1660) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Fix initialization of collider state in MPM finite difference mode (newton-physics#1652) * docs: document ModelBuilder.default_shape_cfg (newton-physics#1662) * Finalize the collision API (newton-physics#1581) * Remove hardcoded subnet ID from AWS workflow (newton-physics#1664) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Attempt to fix AWS config (newton-physics#1666) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Update AWS workflows to g7e.2xlarge with multi-AZ failover (newton-physics#1669) * fix(viewer-usd): disambiguate log_points colors for N=3 warp arrays (newton-physics#1661) * Viewer gl optimizations (newton-physics#1656) Signed-off-by: Miles Macklin <mmacklin@nvidia.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: Eric Heiden <eheiden@nvidia.com> * docs: add articulation workflow guidance and regression check (newton-physics#1663) * fix(examples): propagate IK solution to model state in Franka example (newton-physics#1637) * fix(deps,docs): bump nbconvert to 7.17.0 and fix ArticulationView doctest (newton-physics#1670) * Cleanup and improve some example (newton-physics#1625) Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Handle zero-mass bodies and flip ensure_nonstatic_links default (newton-physics#1635) * Additional testing for ArticulationView (newton-physics#1527) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Update warp-lang nightly to 1.12.0.dev20260217 (newton-physics#1677) * Change default friction coefficients to match MuJoCo (newton-physics#1681) * Refactor mesh creation functions (newton-physics#1654) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Parse joint frictionloss from MJCF (newton-physics#1680) * Fix free joint body_pos and add ref/qpos0 support for MuJoCo solver (newton-physics#1645) * Fix root shapes in ArticulationView with fixed base (newton-physics#1639) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Use model collision methods and remove `create_collision_pipeline` from examples (newton-physics#1648) Co-authored-by: nvtw <110816143+nvtw@users.noreply.github.com> * Fix XPBD apply_joint_forces ignoring child joint transform (newton-physics#1582) * Adjust SDF API (newton-physics#1644) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Optimize test suite runtime (~18% faster) (newton-physics#1689) * Remove ensure_nonstatic_links option from importers (newton-physics#1682) Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> * SolverMuJoCo: Add geom_margin support, align thickness default with MuJoCo and schemas (newton-physics#1653) * refactor: privatize non-public solver internals (newton-physics#1683) * Fix option parsing with multiple <option> elements from includes (newton-physics#1692) * Bump warp-lang nightly and newton-usd-schemas (newton-physics#1693) * Get rid of tkinter dependency (newton-physics#1676) * Fix SDF example contact buffer overflow (newton-physics#1695) * Fix implicit biastype for position/velocity actuator shortcuts (newton-physics#1678) * Fix include processor to respect meshdir/texturedir (newton-physics#1685) * Reduce cold-cache Warp compile time for geometry modules (newton-physics#1618) Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> * Fix xyzw-to-wxyz quaternion conversion in body inertia kernel (newton-physics#1694) * Rename key to label and add hierarchical labels (newton-physics#1592) (newton-physics#1632) * Expose geometry SDF helpers on public API (newton-physics#1684) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Improve custom frequency handling from USD, parse MuJoCo actuators and tendons (newton-physics#1510) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Collapse fixed joints with non articulated bodies (newton-physics#1608) * Fix renaming joint_key -> joint_label (newton-physics#1700) Signed-off-by: Eric Heiden <eheiden@nvidia.com> * Bump .python-version from 3.11 to 3.12 (newton-physics#1702) * Replace CITATION.md with CITATION.cff (newton-physics#1706) * Respect MJCF contype=conaffinity=0 via collision_group=0 (newton-physics#1703) * Make ViewerUSD reuse existing USD layers for the same output path (newton-physics#1704) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * Remove dead up-axis conversion from MuJoCo solver (newton-physics#1707) * Skip IK cube stacking example test (newton-physics#1713) * Fix global pairs (world=-1) not exported to MuJoCo spec (newton-physics#1705) * Reduce the memory consumption of hydroelastic contacts (newton-physics#1609) * Fix flakiness cube stacking (newton-physics#1714) * Fix collision shapes not toggleable in viewer UI (newton-physics#1715) * Fix softbody examples table layout in README (newton-physics#1716) * Standardize sensor APIs: label matching, keyword args, and update() method naming (newton-physics#1665) * Fix intermittent crash in parallel test runner from Manager proxy race (newton-physics#1721) * Clean up inertia.py function arguments to match Mesh.create_* API (newton-physics#1719) * Reduce default test runner verbosity (newton-physics#1723) * Add menagerie comparison tests for SolverMuJoCo (newton-physics#1720) Signed-off-by: Alain Denzler <adenzler@nvidia.com> Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> * Add spatial tendon support for MuJoCo solver (newton-physics#1687) Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> * Expose qfrc_actuator from mujoco (newton-physics#1698) * Skip non-MODEL custom attributes in finalize validation (newton-physics#1734) * Resolve inheritrange for position actuators in MJCF parser (newton-physics#1727) * Support dampratio for position/velocity actuator shortcuts (newton-physics#1722) * Limit concurrency to 1 (newton-physics#1736) * Add a helper method for checking applied usd (newton-physics#1731) * Enhance playback URL handling in ViewerViser (newton-physics#1742) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Removed RenderShapeType (newton-physics#1748) * Override only MassAPI attributes that have been authored (newton-physics#1688) * Integration of newton-actuators (newton-physics#1342) Signed-off-by: jvonmuralt <jvonmuralt@nvidia.com> * Fix ArticulationView crash with fixed-joint-only articulations (newton-physics#1726) * Improve retrieval of Jupyter base URL in ViewerViser (newton-physics#1750) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Disable dynamics testing for UR5e and Apollo tests to avoid CI flakiness (newton-physics#1755) * Margin and Gap rename (newton-physics#1732) * Vbd Demos Fixing (newton-physics#1740) * Fix ViewerViser.log_lines method (newton-physics#1764) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Update warp-lang dependency to 1.12.0rc1 (newton-physics#1763) * Fix fromto capsule/cylinder orientation in MJCF parser (newton-physics#1741) * fix: multi-world particle BVH indexing (newton-physics#1641) Co-authored-by: Eric Heiden <eheiden@nvidia.com> * Clean up unused and internal-only kwargs in SolverMuJoCo (newton-physics#1766) * Parsing of the mimic joint and contact gap/margin from newton schemas (newton-physics#1690) Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> * API Refactor v2 (newton-physics#1749) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Support explicit geom mass attributes in MJCF (newton-physics#1744) * Bump flask and werkzeug in lockfile for security (newton-physics#1769) Co-authored-by: Cursor <cursoragent@cursor.com> * Split MJCF worldbody root bodies into separate articulations (newton-physics#1754) * Expose VBD rigid contact forces for solver coupling (newton-physics#1745) Signed-off-by: JC <jumyungc@nvidia.com> * Add MJCF ellipsoid geom import and regression test (newton-physics#1772) Co-authored-by: Cursor <cursoragent@cursor.com> * Weld equality constraints parsed from mjcf are given Nan as the default value of torquescale. The correct default should be 1.0 (newton-physics#1760) * Improve picking accuracy and stability (newton-physics#1712) * Franka cloth demo improvement (newton-physics#1765) * Support computing sensing object transforms & API cleanup (newton-physics#1759) * Remove threading workaround (newton-physics#1751) * [Warp Raytrace] Consolidated ray intersect functions, renamed Options to Config (newton-physics#1767) * Improve README example gallery for PyPI compatibility (newton-physics#1776) * Fix issue with mesh in rerun viewer (newton-physics#1768) * Add PhysxMimicJointAPI parsing to USD importer (newton-physics#1735) * Move some math functions to Warp (newton-physics#1717) Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Add test to ensure MJCF xform argument is relative (newton-physics#1777) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Emit diaginertia instead of fullinertia for diagonal body inertia (newton-physics#1780) * Change default joint armature from 0.01 to 0 (newton-physics#1782) * Fix default kp/kv for position and velocity actuators (newton-physics#1786) * Lock body inertia after explicit MJCF <inertial> element (newton-physics#1784) * Fix for MJCF actuator custom attributes (newton-physics#1783) * Bump version to 0.2.3 Prepare the package metadata for the v0.2.3 release tag. * Fix ViewerRerun rendering of instances from hidden meshes (newton-physics#1788) * API cleanup (newton-physics#1789) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * BODY actuator target name bypasses body name de-duplication in SolverMuJoCo (newton-physics#1729) * Use default density for visual geoms in MJCF import (newton-physics#1781) * Fix GL viewer crash on Wayland (newton-physics#1793) * Make USD xform parameter control absolute articulation placement (newton-physics#1771) * Fix CUDA context corruption in SDF implementation (newton-physics#1792) * Bump mujoco-warp dependency to 3.5.0.2 (newton-physics#1779) * Fix MuJoCo margin/gap conversion (newton-physics#1785) * Bump version to 1.1.0.dev0 (newton-physics#1798) * Missing unittest.main added back to test_import_mjcf.py. Helps with F5 debugging in VS Code. (newton-physics#1796) * Improve H1 example (newton-physics#1801) Signed-off-by: Eric Heiden <eric-heiden@outlook.com> * Fix ViewerViser.set_camera() (newton-physics#1805) * Rename examples to follow prefix-first naming convention (newton-physics#1802) * Improve MuJoCo actuator domain randomization (newton-physics#1773) * Restore in_cup test in hydro example (newton-physics#1775) * Fix `newton.geometry` imports and change of Mesh maxhullverts global constant * Adapt to having both margin and gap arrays for each geom/shape * Fix for newton.geometry API changes in primitive/narrowphase.py * Fix removal of `BroadPhaseMode` IntEnum and revert to newton-sytle string literals * WIP: Adapt geometry/unified.py to fix breaking changes to `newton.geometry` API * First pass of API adjustments * Patch more gaps (with respect to margin and gap renaming) * GeoType.SDF was removed - reflect that in Kamino * Introduce CoM position offsets w.r.t body frame and operations to convert between body CoM state and generic local body-fixed reference frames. * Add caching of per-entity labels/names/keys in the Model subcontainers * Remove SDF shape wrapper since it's now internal to mesh handling in CD pipelines * Fix USD test assets * Add Newton <--> Kamino joint type conversion operations and per-space default limit constants * Add some cleanup to geometry and unified CD + UTs * Add Newton <--> Kamino shape type conversion operations * Apply new default joint coord limit constants to limits.py * Adapt foubrar model builder and USD asset to produce the same result in sim example * Purge "physical" goems and collapse all into a single group, and purge geometry "layers" * Disable allocation per-joint wrenches by default and make them optional * Make `Model` a dataclass * Separate ModelData* containers into own `core/data.py` module * Fix imports of ModelData * Rename `Model` as `ModelKamino` * Rename `ModelData` as `DataKamino` * Rename `State` as `StateKamino` * Rename `Control` as `ControlKamino` * Rename `Limits` as `LimitsKamino` * Rename `Contacts` as `ContactsKamino` * Rename `ModelBuilder` as `ModelBuilderKamino` * Make imports in test utilities relative * Revise CD meta-data attributes and their computation in GeometryModel and ModelBuilderKamino * Revise primitive CD pipeline * Revise unified CD pipeline * Revise CD front-end interfaces * Fix UTs and relevant utils for interface changes to CD * Change to `wp.DeviceLike` to account for upcoming deprecation of `Devicelike` * Depracate legacy HDF5 data io (will be replaced in the future) * Fix banned imports at module level * Modify USD importer to detect articulations and order geoms and joints similarly to how the Newton `parse_usd` function does. * Add conversion operation from `newton.Model` to `ModelKamino` * Add data, state and control container conversions * Add SolverKamino wrapper that fullfils newton integration interface * Add newton integration examples * Add SolverKamino to newton solver module imports * WIP: fix problem with lambda_j being allocated for only kinematic constraints and failing on array copying * Fix banned git import in benchmark * Rename *Settings to *Config (#213) * Rename SolverKaminoSettings -> SolverKaminoConfig * Rename DualProblemSettings -> DualProblemConfig * Rename CollisionDetectorSettings -> CollisionDetectorConfig * Rename PADMMSettings -> PADMMConfig * Rename ForwardKinematicsSolverSettings -> ForwardKinematicsSolverConfig * Rename SimulatorSettings -> SimulatorConfig * Add check for model compatibility in SolverKamino (#209) * Fix device assignment in sparse CG test on CPU (#216) * Replace Enum-type config attributes with Literal (#215) * Replace warmstart mode config param with literal * Replace contact warmstart mode config param with literal * Replace rotation correction config param with literal * Replace penalty update config param with literal * Replace FK preconditioner option config param with literal * Add post-init checks for dual/PADMM configs * Rename FKPreconditionerOptions to FKPreconditionerType * Remove WorldDescriptor from ModelKamino (#219) * Add geom index offset array to model info * Replace access to world description in model * Remove world descriptor from model * Fix computation of kinematics residual with sparse Jacobian (#220) * Migrates `ModelKaminoSize` to it's own module to avoid circular dependency between core/model.py and core/state.py and rename it as `SizeKamino`. * WIP: Fix SolverKamino wrapper * Fix circular dependency in conversions.py * Fix some broken unit tests and WIP to fix fourbar contact conversions and rendering * Make some conversions @staticmethods instead, because they dont need common class attributes * Fix declaration of custom state attributes and their conversion to/from newton and kamino * WIP: Debug model conversion and newton sim examples * Rename and cleanup start index array of per-world geoms * Model conversion and newton sim examples now work. * Make gravity conversion operation re-usable * Migrates boxes_fourbar builder using newton.ModelBuilder to it's own module to prepare for migration. * Use gravity conversion utility func in SolverKamino * Add reusable joint-parameterization conversion utility * Remove world-descriptor from model converter * Rename helper converter that handles entity-local transforms * Add some cleanup to DR Legs, ANYmal D and basic four-bar examples * Fix module-level imports of additional kamino-specific development dependencies * Fix module-level imports of additional kamino-specific development dependencies * Fix erroneous merge conflict. --------- Signed-off-by: Alain Denzler <adenzler@nvidia.com> Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com> Signed-off-by: JC <jumyungc@nvidia.com> Signed-off-by: Milad Rakhsha <mrakhsha@nvidia.com> Signed-off-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Signed-off-by: Eric Heiden <eric-heiden@outlook.com> Signed-off-by: Eric Heiden <eheiden@nvidia.com> Signed-off-by: Miles Macklin <mmacklin@nvidia.com> Signed-off-by: adenzler-nvidia <adenzler@nvidia.com> Signed-off-by: jvonmuralt <jvonmuralt@nvidia.com> Co-authored-by: Daniela Hase <116915287+daniela-hase@users.noreply.github.com> Co-authored-by: adenzler-nvidia <adenzler@nvidia.com> Co-authored-by: Viktor Reutskyy <33062116+vreutskyy@users.noreply.github.com> Co-authored-by: Eric Shi <97630937+shi-eric@users.noreply.github.com> Co-authored-by: Anka Chen <AnkaChan@users.noreply.github.com> Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com> Co-authored-by: JC-nvidia <116605903+jumyungc@users.noreply.github.com> Co-authored-by: Kenny Vilella <kvilella@nvidia.com> Co-authored-by: nvtw <110816143+nvtw@users.noreply.github.com> Co-authored-by: Milad-Rakhsha-NV <167464435+Milad-Rakhsha-NV@users.noreply.github.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: Lennart Röstel <65088822+lenroe@users.noreply.github.com> Co-authored-by: Eric Heiden <eheiden@nvidia.com> Co-authored-by: jvonmuralt <jvonmuralt@nvidia.com> Co-authored-by: camevor <camevor@nvidia.com> Co-authored-by: mzamoramora-nvidia <mzamoramora@nvidia.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Alessandro Roncone <alecive87@gmail.com> Co-authored-by: gdaviet <57617656+gdaviet@users.noreply.github.com> Co-authored-by: Miles Macklin <mmacklin@nvidia.com> Co-authored-by: Gordon Yeoman <gyeomannvidia@users.noreply.github.com> Co-authored-by: Lukasz Wawrzyniak <lwawrzyniak@nvidia.com> Co-authored-by: Eric Heiden <eric-heiden@outlook.com> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Viktor Reutskyy <vreutskyy@nvidia.com> Co-authored-by: Lorenzo Terenzi <lorenzoterenzi96@gmail.com> Co-authored-by: smollerNV <164020096+smollerNV@users.noreply.github.com> Co-authored-by: twidmer <twidmer@nvidia.com> Co-authored-by: Christian Schumacher <christian.schumacher@disney.com> Co-authored-by: Guirec-Maloisel <25688871+Guirec-Maloisel@users.noreply.github.com>
3 tasks
mmacklin
pushed a commit
to mmacklin/newton
that referenced
this pull request
Apr 7, 2026
Signed-off-by: Viktor Reutskyy <vreutskyy@nvidia.com>
3 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Closes #1189
Adds native heightfield geometry support to Newton with full collision pipeline integration, MuJoCo solver support, viewer visualization, and an interactive example.
Heightfield API
Heightfields are always static (zero mass/inertia). Data is normalized in the constructor;
min_z/max_zdefine the world-space height mapping.Changes
Core geometry:
Heightfieldclass with auto-normalization, explicit Z range, andmin_z/max_zvalidationHeightfieldDataWarp struct +get_triangle_from_heightfield_cell()for GPU collision kernelsNative collision (CollisionPipeline):
HeightfieldDatapopulated inModelBuilder.finalize()with concatenated elevation arrayNarrowPhasecounter array for cleaner feature-based layoutMuJoCo solver:
hfieldasset with origin shift formin_zMJCF import:
<hfield>assets (nrow, ncol, size, file, inline elevation)elevationattributeViewer:
create_mesh_heightfield()conversionExample:
basic_heightfield: wave terrain with dropping spheres, supports--solver xpbd(native) and--solver mujocoTests
15 tests covering: