roboticstoolbox-python 1.4.3__tar.gz → 1.4.4__tar.gz

roboticstoolbox_python-1.4.4/.github/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+    ".": "1.4.4"
+}

roboticstoolbox_python-1.4.4/.github/workflows/dependabot-auto-merge.yml

@@ -0,0 +1,25 @@
+name: Dependabot auto-merge
+
+on: pull_request
+
+permissions:
+  contents: write
+  pull-requests: write
+
+jobs:
+  dependabot:
+    runs-on: ubuntu-latest
+    if: ${{ github.actor == 'dependabot[bot]' }}
+    steps:
+      - name: Dependabot metadata
+        id: metadata
+        uses: dependabot/fetch-metadata@v3
+        with:
+          github-token: "${{ secrets.GITHUB_TOKEN }}"
+
+      - name: Enable auto-merge for patch/minor bumps
+        if: ${{ steps.metadata.outputs.update-type == 'version-update:semver-patch' || steps.metadata.outputs.update-type == 'version-update:semver-minor' }}
+        run: gh pr merge --auto --merge "$PR_URL"
+        env:
+          PR_URL: ${{ github.event.pull_request.html_url }}
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/CHANGELOG.md

@@ -1,5 +1,75 @@
 # Changelog
 
+## Highlights
+
+**Correctness fix in inverse dynamics (RNE).** `Robot.rne()` — used by every
+URDF/ETS-based robot's inverse dynamics — was silently returning wrong
+torques for any robot with translational offsets between links (i.e.
+virtually all real robots), traced to an upstream `spatialmath-python`
+change to how spatial forces transform. Fixed, along with two related bugs
+in how static (non-actuated) link mass gets attributed to the correct
+joint. Verified against the compiled C++ reference path and textbook
+closed-form solutions.
+
+**Tool transforms now work everywhere.** `robot.tool`, once set, is
+automatically honored by `fkine`/`jacob0`/`jacobe`/`hessian0`/`hessiane`
+(previously silently ignored unless passed explicitly each call), and all
+seven IK solvers (`ik_LM`/`ik_NR`/`ik_GN`/`ikine_LM`/`ikine_NR`/`ikine_GN`/
+`ikine_QP`) now accept a `tool=` argument directly.
+
+**New:** `Panda(use_robot_descriptions=True)` — an opt-in alternate Panda
+model with real inertial data, for users who need correct dynamics rather
+than the default's faster collision geometry.
+
+**Bug fixes:** `joint_velocity_damper` keyword-argument misuse across
+several example scripts; `DHRobot.ikine_6s` undoing the tool transform on
+the wrong side; UR5's gripper link now resolved by name instead of a
+fragile raw index.
+
+**Wiki refresh (not reflected in this changelog — wiki edits are direct
+pushes to a separate repo, not commits here):** a full currency pass across
+all 26 pages, stale/incorrect content fixed, ROS references removed, and
+new pages added — Panda model comparison, reactive control, inverse
+kinematics, and IK/RNE benchmark results with real numbers.
+
+
+## [1.4.4](https://github.com/petercorke/robotics-toolbox-python/compare/v1.4.3...v1.4.4) (2026-09-20)
+
+
+### Features
+
+* **urdf:** Panda(use_robot_descriptions=True) option; doc the tradeoff ([#687](https://github.com/petercorke/robotics-toolbox-python/issues/687)) ([124882b](https://github.com/petercorke/robotics-toolbox-python/commit/124882b06880dff49c015e1156041435eaa0023c))
+
+
+### Bug Fixes
+
+* **dhrobot:** ikine_6s undoes the tool transform on the wrong side ([#680](https://github.com/petercorke/robotics-toolbox-python/issues/680)) ([9fbffc6](https://github.com/petercorke/robotics-toolbox-python/commit/9fbffc6f09265f084e1b53c04b09a75a4aced643)), closes [#640](https://github.com/petercorke/robotics-toolbox-python/issues/640)
+* **docs:** mock bdsim imports so blocks pages render in docs CI ([436e1f3](https://github.com/petercorke/robotics-toolbox-python/commit/436e1f3bbd26faa26eed97ddd47cbf47609bf216))
+* **examples:** call joint_velocity_damper with keyword args ([#690](https://github.com/petercorke/robotics-toolbox-python/issues/690)) ([95ac98e](https://github.com/petercorke/robotics-toolbox-python/commit/95ac98e31876b2226b9f51778ba149e105de78c7))
+* **examples:** call joint_velocity_damper with keyword args in fetch_vision.py ([#695](https://github.com/petercorke/robotics-toolbox-python/issues/695)) ([7aa2d31](https://github.com/petercorke/robotics-toolbox-python/commit/7aa2d317445c84e8eeb8ea64e78cafce2fb938d5))
+* **joint_velocity_damper:** fix bug in q usage ([#666](https://github.com/petercorke/robotics-toolbox-python/issues/666)) ([8246559](https://github.com/petercorke/robotics-toolbox-python/commit/8246559202dd1722f8b6538daad66aed735ca928))
+* **kinematics:** self.tool now applies by default; ik_LM/ik_NR/ik_GN/ikine_XX gain a tool= kwarg ([#697](https://github.com/petercorke/robotics-toolbox-python/issues/697)) ([61d078f](https://github.com/petercorke/robotics-toolbox-python/commit/61d078fcdf419199209ef5cf931c2c2ef51e0859))
+* link_collision_damper ([#685](https://github.com/petercorke/robotics-toolbox-python/issues/685)) ([c040522](https://github.com/petercorke/robotics-toolbox-python/commit/c04052208094fe458efd707c84fe970580ac2fe2))
+* preserve joint types and limits in numerical IK ([e122ccb](https://github.com/petercorke/robotics-toolbox-python/commit/e122ccb9a1ad003879a63ff0fa6d3accdcddde26))
+* **rne:** attribute a sandwiched static link's mass to its preceding joint, not the following one ([#693](https://github.com/petercorke/robotics-toolbox-python/issues/693)) ([bebede0](https://github.com/petercorke/robotics-toolbox-python/commit/bebede019601f4b641cdc5ed8e390f480fad766d))
+* **rne:** don't crash when a joint's kinematic parent is a base-mounted static link ([#696](https://github.com/petercorke/robotics-toolbox-python/issues/696)) ([dba179e](https://github.com/petercorke/robotics-toolbox-python/commit/dba179ed84eae8162257d434197d77609ceb0b15))
+* **rne:** propagate backward-pass force through the correct transform direction ([#691](https://github.com/petercorke/robotics-toolbox-python/issues/691)) ([a80c9c1](https://github.com/petercorke/robotics-toolbox-python/commit/a80c9c1c4c609c35494d7f127054c408f993635f))
+* **rne:** trailing static links after the last joint had their mass dropped ([#684](https://github.com/petercorke/robotics-toolbox-python/issues/684)) ([0d76187](https://github.com/petercorke/robotics-toolbox-python/commit/0d76187b21d86713fc13a36ed6bc00a275eab858))
+* **rtbtool:** drop stray quote in banner's import line ([#678](https://github.com/petercorke/robotics-toolbox-python/issues/678)) ([b36c398](https://github.com/petercorke/robotics-toolbox-python/commit/b36c398aa81de96bc0784c25601ade1bb95e7e73))
+* **urdf:** resolve UR5's gripper link by name, not a raw index ([#683](https://github.com/petercorke/robotics-toolbox-python/issues/683)) ([12ba598](https://github.com/petercorke/robotics-toolbox-python/commit/12ba598a58c1ae21db9145ce4dbca798e8065652)), closes [#578](https://github.com/petercorke/robotics-toolbox-python/issues/578)
+
+
+### Documentation
+
+* **models:** credit robot_descriptions in the 8 model classes that use it ([#681](https://github.com/petercorke/robotics-toolbox-python/issues/681)) ([5dc8618](https://github.com/petercorke/robotics-toolbox-python/commit/5dc86184df70ec4cc639db76d203fb534ce8a3f7))
+* **models:** note UR10's near-singular inertia matrix is manufacturer data ([#682](https://github.com/petercorke/robotics-toolbox-python/issues/682)) ([97b0a3f](https://github.com/petercorke/robotics-toolbox-python/commit/97b0a3fa554733ca1247793f199cc6a6d9ba40c1))
+* **readme:** move research applications section to the wiki ([#689](https://github.com/petercorke/robotics-toolbox-python/issues/689)) ([8facc5e](https://github.com/petercorke/robotics-toolbox-python/commit/8facc5eeef8fc5c9ec93b5f1c1f340b0a4fa03ea))
+
+
+### Miscellaneous
+
+* release as 1.4.4 ([758a81b](https://github.com/petercorke/robotics-toolbox-python/commit/758a81b26fe1302a660aeba3a47b79abb76c2855))
+
 ## [1.4.3](https://github.com/petercorke/robotics-toolbox-python/compare/v1.4.2...v1.4.3) (2026-09-12)
 
 

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: roboticstoolbox-python
-Version: 1.4.3
+Version: 1.4.4
 Summary: A Python library for robotics education and research
 Keywords: python,robotics,robotics-toolbox,kinematics,dynamics,motion-planning,trajectory-generation,jacobian,hessian,control,simulation,robot-manipulator,mobile-robot
 Author-Email: Jesse Haviland <j.haviland@qut.edu.au>, Peter Corke <rvc@petercorke.com>
@@ -39,7 +39,7 @@ Project-URL: documentation, https://petercorke.github.io/robotics-toolbox-python
 Project-URL: repository, https://github.com/petercorke/robotics-toolbox-python
 Requires-Python: >=3.10
 Requires-Dist: numpy
-Requires-Dist: spatialmath-python>=1.1.16
+Requires-Dist: spatialmath-python>=1.1.18
 Requires-Dist: spatialgeometry>=1.4.0
 Requires-Dist: pgraph-python
 Requires-Dist: scipy
@@ -543,44 +543,4 @@ Graphical visualisation via Swift is currently not supported under Windows. Howe
 
 The toolbox is incredibly useful for developing and prototyping algorithms for research, thanks to the exhaustive set of well documented and mature robotic functions exposed through clean and painless APIs. Additionally, the ease at which a user can visualize their algorithm supports a rapid prototyping paradigm.
 
-### Publication List
-
-J. Haviland, N. Sünderhauf and P. Corke, "**A Holistic Approach to Reactive Mobile Manipulation**," in _IEEE Robotics and Automation Letters_, doi: 10.1109/LRA.2022.3146554. In the video, the robot is controlled using the Robotics toolbox for Python and features a recording from the [Swift](https://github.com/jhavl/swift) Simulator.
-
-[[Arxiv Paper](https://arxiv.org/abs/2109.04749)] [[IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/9695298)] [[Project Website](https://jhavl.github.io/holistic/)] [[Video](https://youtu.be/-DXBQPeLIV4)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/holistic_mm_non_holonomic.py)]
-
-<p>
-  <a href="https://youtu.be/-DXBQPeLIV4">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/holistic_youtube.png" width="560">
-  </a>
-</p>
-
-J. Haviland and P. Corke, "**NEO: A Novel Expeditious Optimisation Algorithm for Reactive Motion Control of Manipulators**," in _IEEE Robotics and Automation Letters_, doi: 10.1109/LRA.2021.3056060. In the video, the robot is controlled using the Robotics toolbox for Python and features a recording from the [Swift](https://github.com/jhavl/swift) Simulator.
-
-[[Arxiv Paper](https://arxiv.org/abs/2010.08686)] [[IEEE Xplore](https://ieeexplore.ieee.org/document/9343718)] [[Project Website](https://jhavl.github.io/neo/)] [[Video](https://youtu.be/jSLPJBr8QTY)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/neo.py)]
-
-<p>
-  <a href="https://youtu.be/jSLPJBr8QTY">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/neo_youtube.png" width="560">
-  </a>
-</p>
-
-K. He, R. Newbury, T. Tran, J. Haviland, B. Burgess-Limerick, D. Kulić, P. Corke, A. Cosgun, "**Visibility Maximization Controller for Robotic Manipulation**", in _IEEE Robotics and Automation Letters_, doi: 10.1109/LRA.2022.3188430. In the video, the robot is controlled using the Robotics toolbox for Python and features a recording from the [Swift](https://github.com/jhavl/swift) Simulator.
-
-[[Arxiv Paper](https://arxiv.org/abs/2202.12557)] [[IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/9815144)] [[Project Website](https://rhys-newbury.github.io/projects/vmc/)] [[Video](https://youtu.be/vobLvg4E3kM)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/fetch_vision.py)]
-
-<p>
-  <a href="https://youtu.be/vobLvg4E3kM">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/vmc_youtube.png" width="560">
-  </a>
-</p>
-
-**A Purely-Reactive Manipulability-Maximising Motion Controller**, J. Haviland and P. Corke. In the video, the robot is controlled using the Robotics toolbox for Python.
-
-[[Paper](https://arxiv.org/abs/2002.11901)] [[Project Website](https://jhavl.github.io/mmc/)] [[Video](https://youtu.be/Vu_rcPlaADI)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/mmc.py)]
-
-<p>
-  <a href="https://youtu.be/Vu_rcPlaADI">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/mmc_youtube.png" width="560">
-  </a>
-</p>
+It's been the implementation substrate for several published reactive-control papers, and its kinematics are a direct implementation of a two-part differential-kinematics tutorial — see the wiki's [Reactive control](https://github.com/petercorke/robotics-toolbox-python/wiki/Reactive-control) and [Inverse kinematics](https://github.com/petercorke/robotics-toolbox-python/wiki/Inverse-kinematics) pages.

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/README.md

@@ -443,44 +443,4 @@ Graphical visualisation via Swift is currently not supported under Windows. Howe
 
 The toolbox is incredibly useful for developing and prototyping algorithms for research, thanks to the exhaustive set of well documented and mature robotic functions exposed through clean and painless APIs. Additionally, the ease at which a user can visualize their algorithm supports a rapid prototyping paradigm.
 
-### Publication List
-
-J. Haviland, N. Sünderhauf and P. Corke, "**A Holistic Approach to Reactive Mobile Manipulation**," in _IEEE Robotics and Automation Letters_, doi: 10.1109/LRA.2022.3146554. In the video, the robot is controlled using the Robotics toolbox for Python and features a recording from the [Swift](https://github.com/jhavl/swift) Simulator.
-
-[[Arxiv Paper](https://arxiv.org/abs/2109.04749)] [[IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/9695298)] [[Project Website](https://jhavl.github.io/holistic/)] [[Video](https://youtu.be/-DXBQPeLIV4)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/holistic_mm_non_holonomic.py)]
-
-<p>
-  <a href="https://youtu.be/-DXBQPeLIV4">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/holistic_youtube.png" width="560">
-  </a>
-</p>
-
-J. Haviland and P. Corke, "**NEO: A Novel Expeditious Optimisation Algorithm for Reactive Motion Control of Manipulators**," in _IEEE Robotics and Automation Letters_, doi: 10.1109/LRA.2021.3056060. In the video, the robot is controlled using the Robotics toolbox for Python and features a recording from the [Swift](https://github.com/jhavl/swift) Simulator.
-
-[[Arxiv Paper](https://arxiv.org/abs/2010.08686)] [[IEEE Xplore](https://ieeexplore.ieee.org/document/9343718)] [[Project Website](https://jhavl.github.io/neo/)] [[Video](https://youtu.be/jSLPJBr8QTY)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/neo.py)]
-
-<p>
-  <a href="https://youtu.be/jSLPJBr8QTY">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/neo_youtube.png" width="560">
-  </a>
-</p>
-
-K. He, R. Newbury, T. Tran, J. Haviland, B. Burgess-Limerick, D. Kulić, P. Corke, A. Cosgun, "**Visibility Maximization Controller for Robotic Manipulation**", in _IEEE Robotics and Automation Letters_, doi: 10.1109/LRA.2022.3188430. In the video, the robot is controlled using the Robotics toolbox for Python and features a recording from the [Swift](https://github.com/jhavl/swift) Simulator.
-
-[[Arxiv Paper](https://arxiv.org/abs/2202.12557)] [[IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/9815144)] [[Project Website](https://rhys-newbury.github.io/projects/vmc/)] [[Video](https://youtu.be/vobLvg4E3kM)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/fetch_vision.py)]
-
-<p>
-  <a href="https://youtu.be/vobLvg4E3kM">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/vmc_youtube.png" width="560">
-  </a>
-</p>
-
-**A Purely-Reactive Manipulability-Maximising Motion Controller**, J. Haviland and P. Corke. In the video, the robot is controlled using the Robotics toolbox for Python.
-
-[[Paper](https://arxiv.org/abs/2002.11901)] [[Project Website](https://jhavl.github.io/mmc/)] [[Video](https://youtu.be/Vu_rcPlaADI)] [[Code Example](https://github.com/petercorke/robotics-toolbox-python/blob/main/roboticstoolbox/examples/mmc.py)]
-
-<p>
-  <a href="https://youtu.be/Vu_rcPlaADI">
-    <img src="https://raw.githubusercontent.com/petercorke/robotics-toolbox-python/main/docs/figs/mmc_youtube.png" width="560">
-  </a>
-</p>
+It's been the implementation substrate for several published reactive-control papers, and its kinematics are a direct implementation of a two-part differential-kinematics tutorial — see the wiki's [Reactive control](https://github.com/petercorke/robotics-toolbox-python/wiki/Reactive-control) and [Inverse kinematics](https://github.com/petercorke/robotics-toolbox-python/wiki/Inverse-kinematics) pages.

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/docs/source/IK/ik.rst

@@ -60,6 +60,14 @@ For example when using a 3 DOF manipulator tool orientation might be unimportant
 
 setting ``joint_limits = True`` will reject solutions with joint limit violations. Note that finding a solution with valid joint coordinates is likely to take longer than without.
 
+After convergence, prismatic joint coordinates are left unchanged. Revolute
+coordinates already within their limits are preserved, including limits outside
+the principal interval of :math:`[-\pi, \pi]`. Other revolute coordinates are
+reduced to the principal angle and, if necessary, shifted by whole turns towards
+their limits. If no equivalent angle is within the limits, the solution is
+rejected when ``joint_limits = True``. Setting ``joint_limits = False`` disables
+this rejection; it does not change a prismatic distance into an angle.
+
 .. rubric:: Others
 
 There are other arguments which may be unique to the solver, so check the documentation of the solver you wish to use for a complete list and explanation of arguments.

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/docs/source/conf.py

@@ -63,6 +63,14 @@ bibtex_bibfiles = ["refs.bib"]
 autosummary_generate = True
 autodoc_member_order = "bysource"
 
+# roboticstoolbox.blocks.* subclasses bdsim's block types, but bdsim is a
+# `dev`-only dependency here (unlike MVTB, where it's genuinely optional) --
+# the docs CI job installs `.[docs]` only, so bdsim is never actually
+# importable there. Mock it so autodoc can still introspect the blocks
+# modules' own classes/docstrings without a real bdsim install. Same fix as
+# MVTB's conf.py, which hit this first.
+autodoc_mock_imports = ["bdsim", "bdsim.block_types"]
+
 # Add any paths that contain templates here, relative to this directory.
 templates_path = ["_templates"]
 

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/_cpu_info.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python
-"""Shared helper for the benchmark scripts in this directory (rne_speed.py,
-ik_speed.py) -- factored out once a second script needed the same thing.
+"""Shared helper for the benchmark scripts in this directory (benchmark_rne.py,
+benchmark_ik.py) -- factored out once a second script needed the same thing.
 """
 
 import os

roboticstoolbox_python-1.4.3/examples/ik_speed.py → roboticstoolbox_python-1.4.4/examples/benchmark_ik.py

@@ -23,13 +23,18 @@ ets = robot.ets()
 
 ### Experiment parameters
 # Number of problems to solve for the C-accelerated columns (ik_XX and
-# ikine_XX with the C++ ETS)
-nproblems = 10_000
+# ikine_XX with the C++ ETS). Sized for a ~1-2 minute total run, not for
+# per-solver statistical precision: LM Sugihara converges far slower than
+# the other methods on random targets (tens of ms/problem even with the
+# C++ ETS, ~500x LM Chan's cost) and dominates this phase's total time, so
+# nproblems is capped by its cost rather than the cheaper solvers'.
+nproblems = 1_000
 
 # The pure-Python ETS fallback is interpreted at every iteration, not just
-# the outer solver loop -- 10,000 problems would take far too long, so use
-# a much smaller sample for that column alone.
-nproblems_slow = 1_000
+# the outer solver loop -- full nproblems would take far too long (LM
+# Sugihara alone runs ~2s/problem here), so use a much smaller sample for
+# that column alone. This is a rough comparison, not a precise one.
+nproblems_slow = 15
 
 # Cartesion DoF priority matrix
 mask = np.array([1.0, 1.0, 1.0, 1.0, 1.0, 1.0])

roboticstoolbox_python-1.4.3/examples/rne_speed.py → roboticstoolbox_python-1.4.4/examples/benchmark_rne.py

@@ -19,7 +19,7 @@ the whole trajectory inside a single C++ call rather than once per row --
 see rne.md plan step 7 and tests/test_fknm_fallback.py's
 TestRneTrajectoryVaryingRows for the correctness side of this).
 
-Riffs off ik_speed.py's ANSITable reporting and rne_compare.py's timing
+Riffs off benchmark_ik.py's ANSITable reporting and rne_compare.py's timing
 harness.
 """
 

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/fetch_vision.py

@@ -134,14 +134,16 @@ def step():
 
     # Form the joint limit velocity damper
     Ain[: fetch.n, : fetch.n], bin[: fetch.n] = fetch.joint_velocity_damper(
-        ps, pi, fetch.n
+        ps=ps, pi=pi, n=fetch.n
     )
 
-    Ain_torso, bin_torso = fetch_camera.joint_velocity_damper(0.0, 0.05, fetch_camera.n)
+    Ain_torso, bin_torso = fetch_camera.joint_velocity_damper(
+        ps=0.0, pi=0.05, n=fetch_camera.n
+    )
     Ain[2, 2] = Ain_torso[2, 2]
     bin[2] = bin_torso[2]
 
-    Ain_cam, bin_cam = fetch_camera.joint_velocity_damper(ps, pi, fetch_camera.n)
+    Ain_cam, bin_cam = fetch_camera.joint_velocity_damper(ps=ps, pi=pi, n=fetch_camera.n)
     Ain[n_base + n_arm : n, n_base + n_arm : n] = Ain_cam[3:, 3:]
     bin[n_base + n_arm : n] = bin_cam[3:]
 

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/holistic_mm_non_holonomic.py

@@ -55,7 +55,7 @@ def step_robot(r: rtb.ERobot, Tep):
     pi = 0.9
 
     # Form the joint limit velocity damper
-    Ain[: r.n, : r.n], bin[: r.n] = r.joint_velocity_damper(ps, pi, r.n)
+    Ain[: r.n, : r.n], bin[: r.n] = r.joint_velocity_damper(ps=ps, pi=pi, n=r.n)
 
     # Linear component of objective function: the manipulability Jacobian
     c = np.concatenate(

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/holistic_mm_omni.py

@@ -55,7 +55,7 @@ def step_robot(r: rtb.ERobot, Tep):
     pi = 0.9
 
     # Form the joint limit velocity damper
-    Ain[: r.n, : r.n], bin[: r.n] = r.joint_velocity_damper(ps, pi, r.n)
+    Ain[: r.n, : r.n], bin[: r.n] = r.joint_velocity_damper(ps=ps, pi=pi, n=r.n)
 
     # Linear component of objective function: the manipulability Jacobian
     c = np.concatenate(

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/mmc.py

@@ -79,7 +79,7 @@ while not arrived:
     pi = 0.9
 
     # Form the joint limit velocity damper
-    Ain[:n, :n], bin[:n] = panda.joint_velocity_damper(ps, pi, n)
+    Ain[:n, :n], bin[:n] = panda.joint_velocity_damper(ps=ps, pi=pi, n=n)
 
     # Linear component of objective function: the manipulability Jacobian
     c = np.r_[-panda.jacobm().reshape((n,)), np.zeros(6)]

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/neo.py

@@ -90,7 +90,7 @@ def step():
     pi = 0.9
 
     # Form the joint limit velocity damper
-    Ain[:n, :n], bin[:n] = panda.joint_velocity_damper(ps, pi, n)
+    Ain[:n, :n], bin[:n] = panda.joint_velocity_damper(ps=ps, pi=pi, n=n)
 
     # For each collision in the scene
     for collision in collisions:

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/rne_compare.py

@@ -13,7 +13,7 @@ ETS segments -- true for MDH DHRobots, false for Puma560 (standard DH) --
 and now asserts on the incompatible case instead of silently returning a
 wrong answer. So Robot.rne(puma, ...) below is *expected* to raise, and
 this script demonstrates that rather than crashing on it. For a 3-way
-comparison on a robot where all three genuinely apply, see rne_speed.py
+comparison on a robot where all three genuinely apply, see benchmark_rne.py
 (uses rtb.models.DH.Panda(), which is mdh=True).
 """
 
@@ -73,7 +73,7 @@ t_py = bench("pure Python (rne_python)", lambda: puma.rne_python(Q, QD, QDD))
 print()
 print(f"speedup rne_python -> C:  {t_py/t_c:6.1f}x")
 print("(Robot.rne skipped here -- Puma560 is standard DH, see note above."
-      " For a full 3-way timing comparison, see rne_speed.py.)")
+      " For a full 3-way timing comparison, see benchmark_rne.py.)")
 
 print()
 print("Per-row C-call count check (does rne() batch the trajectory into one")

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/examples/swift_recording.py

@@ -98,7 +98,7 @@ while not arrived:
     pi = 0.9
 
     # Form the joint limit velocity damper
-    Ain[:n, :n], bin[:n] = panda.joint_velocity_damper(ps, pi, n)
+    Ain[:n, :n], bin[:n] = panda.joint_velocity_damper(ps=ps, pi=pi, n=n)
 
     # Linear component of objective function: the manipulability Jacobian
     c = np.r_[-panda.jacobm().reshape((n,)), np.zeros(6)]

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/pyproject.toml

@@ -4,7 +4,7 @@ name = "roboticstoolbox-python"
 
 description = "A Python library for robotics education and research"
 
-version = "1.4.3"
+version = "1.4.4"
 
 authors = [
     { name = "Jesse Haviland", email = "j.haviland@qut.edu.au" },
@@ -13,7 +13,7 @@ authors = [
 
 dependencies = [
     "numpy",
-    "spatialmath-python>=1.1.16",
+    "spatialmath-python>=1.1.18",
     "spatialgeometry>=1.4.0",
     "pgraph-python",
     "scipy",

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/bin/rtbtool.py

@@ -202,7 +202,7 @@ def make_banner():
     from spatialmath import *
     from spatialmath.base import *
     from spatialmath.base import sym
-    from roboticstoolbox import *"
+    from roboticstoolbox import *
 
     # useful variables
     from math import pi

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/ets/cpp-extensions/ik.cpp

@@ -51,8 +51,52 @@ static void _IK_loop(
 
             if (*E < tol)
             {
-                for (int i = 0; i < ets->n; i++)
-                    q(i) = std::fmod(q(i) + PI, PI_x2) - PI;
+                // Preserve translations and coordinates already within their limits.
+                // Other revolute coordinates may move only by whole turns, so the
+                // converged end-effector pose is unchanged (see IKSolver._normalise_q).
+                int j = 0;
+                for (int i = 0; i < ets->m; i++)
+                {
+                    ET *et = ets->ets[i];
+                    if (!et->isjoint)
+                        continue;
+
+                    double lower = ets->qlim_l[j];
+                    double upper = ets->qlim_h[j];
+                    if (et->axis < 3 && !(lower <= q(j) && q(j) <= upper))
+                    {
+                        // Avoid rounding a principal angle across a nearby limit.
+                        double angle = q(j);
+                        if (!(angle >= -PI && angle < PI))
+                        {
+                            // Unlike Python's %, fmod can return a negative remainder.
+                            angle = std::fmod(angle + PI, PI_x2);
+                            if (angle < 0)
+                                angle += PI_x2;
+                            angle -= PI;
+                        }
+                        if (angle < lower)
+                            angle += PI_x2 * std::ceil((lower - angle) / PI_x2);
+                        else if (angle > upper)
+                            angle -= PI_x2 * std::ceil((angle - upper) / PI_x2);
+                        if (!(lower <= angle && angle <= upper))
+                        {
+                            // Recover rounded endpoints only if whole turns
+                            // reconstruct the original coordinate exactly.
+                            for (double bound : {lower, upper})
+                            {
+                                double turns = std::round((q(j) - bound) / PI_x2);
+                                if (turns != 0 && bound + turns * PI_x2 == q(j))
+                                {
+                                    angle = bound;
+                                    break;
+                                }
+                            }
+                        }
+                        q(j) = angle;
+                    }
+                    j++;
+                }
                 *solution = reject_jl ? _check_lim(ets, q) : 1;
                 break;
             }
@@ -315,4 +359,4 @@ extern "C"
         // return q;
     }
 
-} /* extern "C" */
+} /* extern "C" */

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/models/DH/UR10.py

@@ -27,6 +27,11 @@ class UR10(DHRobot):
 
     .. note::
         - SI units are used.
+        - The last link's inertia tensor has :math:`I_{xx} = 0` -- that's not
+          a bug in this model, it's what the manufacturer's own reference
+          (below) publishes. It makes the robot's inertia matrix near-singular
+          (``det(robot.inertia(q)) ~ 2e-6``), so dynamics work that needs a
+          well-conditioned inertia matrix should account for this.
 
     :References:
 

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/models/URDF/Fetch.py

@@ -47,6 +47,9 @@ class Fetch(URDFRobot):
     from a URDF file.  The model describes its kinematic and graphical
     characteristics.
 
+    The model is loaded via the `robot_descriptions
+    <https://github.com/robot-descriptions/robot_descriptions.py>`_ package.
+
     .. runblock:: pycon
 
         >>> import roboticstoolbox as rtb

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/models/URDF/Frankie.py

@@ -13,6 +13,9 @@ class Frankie(URDFRobot):
     from a URDF file.  The model describes its kinematic and graphical
     characteristics.
 
+    The model is loaded via the `robot_descriptions
+    <https://github.com/robot-descriptions/robot_descriptions.py>`_ package.
+
     .. runblock:: pycon
 
         >>> import roboticstoolbox as rtb
@@ -23,8 +26,14 @@ class Frankie(URDFRobot):
 
     - qz, zero joint angle configuration, 'L' shaped configuration
     - qr, vertical 'READY' configuration
-    - qs, arm is stretched out in the x-direction
-    - qn, arm is at a nominal non-singular configuration
+
+    Loaded via the `robot_descriptions
+    <https://github.com/robot-descriptions/robot_descriptions.py>`_ package
+    (real inertial data, plain-mesh collision geometry), unlike
+    :class:`~roboticstoolbox.models.URDF.Panda.Panda`'s bundled default. See
+    the wiki's `Panda models
+    <https://github.com/petercorke/robotics-toolbox-python/wiki/Panda-models>`_
+    page for why the two differ.
 
     .. codeauthor:: Jesse Haviland
     .. sectionauthor:: Peter Corke

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/models/URDF/FrankieOmni.py

@@ -27,8 +27,16 @@ class FrankieOmni(Robot):
 
     - qz, zero joint angle configuration, 'L' shaped configuration
     - qr, vertical 'READY' configuration
-    - qs, arm is stretched out in the x-direction
-    - qn, arm is at a nominal non-singular configuration
+
+    The arm is loaded from the same bundled ``qut_frankie_description`` xacro
+    as :class:`~roboticstoolbox.models.URDF.Panda.Panda`'s default (no
+    inertial data), spliced by hand onto a separately-loaded
+    ``clearpath_ridgeback_description`` mobile base -- it has no
+    ``use_robot_descriptions``-style option, since the base+arm splice is
+    done manually rather than through :class:`URDFRobot`'s own loader. See
+    the wiki's `Panda models
+    <https://github.com/petercorke/robotics-toolbox-python/wiki/Panda-models>`_
+    page for the full picture.
 
     .. codeauthor:: Jesse Haviland
     .. sectionauthor:: Peter Corke

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/models/URDF/Jaco.py

@@ -12,6 +12,9 @@ class Jaco(URDFRobot):
     from a URDF file.  The model describes its kinematic and graphical
     characteristics.
 
+    The model is loaded via the `robot_descriptions
+    <https://github.com/robot-descriptions/robot_descriptions.py>`_ package.
+
     .. runblock:: pycon
 
         >>> import roboticstoolbox as rtb

{roboticstoolbox_python-1.4.3 → roboticstoolbox_python-1.4.4}/src/roboticstoolbox/models/URDF/PR2.py

@@ -12,6 +12,9 @@ class PR2(URDFRobot):
     from a URDF file.  The model describes its kinematic and graphical
     characteristics.
 
+    The model is loaded via the `robot_descriptions
+    <https://github.com/robot-descriptions/robot_descriptions.py>`_ package.
+
     .. runblock:: pycon
 
         >>> import roboticstoolbox as rtb