Diego Ferigo

A MWE of a `Time` class should at least have the following features: - Implemented as a PyTree (therefore, supporting `jax.jit`). - Have a resolution of 1-10 nanoseconds, and at...

This is a good idea, I'll have a look at [`functools.total_ordering`](https://docs.python.org/3.10/library/functools.html#functools.total_ordering) for the comparison operators. However, currently the main difficulty is in handling properly the overflow when JAX runs in...

It would be nice also adding a new `VelRepr.Centroidal` to obtain the centroidal dynamics using the frame $G := ({}^W \mathbf{p}_{\text{CoM}}, \, [W])$ as base frame. However, this change is...

Many quantities related to centroidal dynamics have been implemented in #117.

The last interesting quantity related to the CoM has been implemented in #161. I think that now JaxSim provides the same quantities as iDynTree. This issue can be closed as...

In these days I realized that any option that replaces a single float with a class to handle the simulation time would introduce major challenges in applications that need to...

JaxSim finally gained official support of 32-bit precision and TPU hardware accelerators: - https://github.com/ami-iit/jaxsim/pull/262 - https://github.com/ami-iit/jaxsim/pull/263

Some questions. When I checked your code a while ago, if I recall it was missing the base-related blocks of the matrix. Is this the case? And I mean, given...

For the records, for those that are not familiar with the `high_level` resources, the left-hand side of the previous equation is no more no less than [`Model.free_floating_bias_forces`](https://github.com/ami-iit/jaxsim/blob/ed05df47a0484280d478f33712e34a69ccb67d4c/src/jaxsim/high_level/model.py#L867-L878).

Waiting for the support of computing directly $\dot{J}$, that would enable the computation of $\dot{M}(\mathbf{q})$ and as a consequence $C(\mathbf{q}, \boldsymbol{\nu})$ (#41), I implemented $\dot{J} \boldsymbol{\nu}$ in #127. This should...