Sudeep, FeatherstoneMethod should not require any implementation of joints, I think it should work with knowledge of the rigid bodies and particles in a system and their relative kinematics. That would keep things general. I do think that a nice implementation of a spatial vector would provide the best foundation for building the Featherstone algorithm and allow it to work with the classes we have already established for describing the kinematics of a system.
Jason moorepants.info +01 530-601-9791 On Mon, Feb 8, 2021 at 9:08 AM Sudeep Sidhu <sudeepmanilsi...@gmail.com> wrote: > Jason, > > I would like to work in this realm in this year's GSoC. > > I went through Sahil Shekhewat's and James Milam's (jbm950) previous > unmerged GSoC work and found out that(I may be wrong) *FeatherStoneMethod > *can't be added until we add fully working *JointsMethod* class and > implement *spatial vectors*. > > I would like to implement *Spatial Vectors* in this year's GSoC and would > like to discuss it further. > > Sudeep Sidhu > > > > >> On Sun, Feb 7, 2021 at 2:24 PM Jason Moore <moore...@gmail.com> wrote: >> >>> Sudeep, >>> >>> The topics related to sympy.physics.vector/mechanics are still >>> possibilities. I will have time to mentor this summer if someone wants to >>> do projects in this realm. >>> >>> We have not updated the ideas page yet for this year so those could be >>> adjusted. Off the top of my head here are some things that I think are >>> priorities: >>> >>> - Finish and enhance the work of Sahil Shekhewat so that models can be >>> built with body and joint specifications (unmerged GSOC work). >>> - Finish and enhance the work of James Milam (jbm950) that adds a >>> FeatherstoneMethod. This is one way to increase the computational >>> efficiency. One thing that is missing are nice implementations of spatial >>> vectors and their operators. >>> - Finish and enhance the work of Nikhil Pappu. The Autolev parser needs >>> to be battle tested on some examples and bugs worked out. We need the tests >>> in the private gitlab repo to actually be run by SymPy. (merged, but not >>> polished GSOC work). >>> - The Linearizer class was updated by James Crist, but I think it is >>> effectively broken for more complex problems. This needs to be fixed and we >>> need examples of it working for systems with holonomic and nonholonomic >>> constraints. >>> - Improve symbolic computational speed. Hard examples need to be >>> profiled and the Python implementations improved, work on the core >>> differentiation algorithms to maximize speed, and ensure that optional >>> dependencies on symengine function and help for hard problems. >>> - Develop a more comprehensive set of examples. I've started creating >>> more and migrating threse to the PyDy documentation: >>> https://pydy.readthedocs.io/en/latest/#examples. One barrier to user >>> adoption is the lack of examples that are clearly written that cover all >>> types of dynamic systems. >>> - I've recently discovered that for some problems the resulting symbolic >>> equations are in a form that results in numerical error accumulation in the >>> arithmetic. This is problematic and figuring out what this issue is and >>> remedying it would be a nice improvement. >>> - All of these PRs are hanging: >>> https://github.com/sympy/sympy/pulls?q=is%3Aopen+is%3Apr+label%3Aphysics.mechanics >>> and it would be nice to resolve them and get them merged. >>> - If work can be done on PyDy, as has in the past, there are several >>> things there too 1) support DAEs, 2) improve the visualizer in a number of >>> ways, 3) migrate examples to jupyter-sphinx, etc. >>> >>> At this point, I'm generally in disfavor of proposing any new features >>> or extensions to the library over fixing and improving what we already >>> have. As you can see, we have several GSoC projects that were not fully >>> polished off or were not merged at all. >>> >>> Jason >>> moorepants.info >>> +01 530-601-9791 <(530)%20601-9791> >>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "sympy" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to sympy+un...@googlegroups.com. >>> >> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/sympy/CAP7f1AgguJTXrTiNpZVS-oe9-jT8BjeSqRiJRPCmbhj4mC4b5A%40mail.gmail.com >>> <https://groups.google.com/d/msgid/sympy/CAP7f1AgguJTXrTiNpZVS-oe9-jT8BjeSqRiJRPCmbhj4mC4b5A%40mail.gmail.com?utm_medium=email&utm_source=footer> >>> . >>> >> -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sympy+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sympy/46e97251-1908-4bc4-af73-20faba189babn%40googlegroups.com > <https://groups.google.com/d/msgid/sympy/46e97251-1908-4bc4-af73-20faba189babn%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAP7f1AirYPVS0vaN-kxmw%2BbqzHfwTA%2BRB0SmP7cU9BEQ-aPGgA%40mail.gmail.com.
