Any function to calculate automatically Qdot and Qddot from an experimental Q (or post optim) ?

[ANaaim]

Hello,

I do not found any function to automatically calculate naturalVelocities and naturalAccelerations from a naturalsegment (from experimental data). Is there any ? I might have not looked properly.

If not might be interesting for beginner user in order to have them only to give the Q extracted in the inverse function. In that case we would calculate ourselve the qdot and qddot.

Best regards

[Ipuch]

Can you show me, how should it look like? i.e. how would you like to use it in bioNC ?

What are the computational steps behind ?

The most close feature I see, is the method _Q_from markers that estimates Q from marker locations in global.

This fonctions are set when building the model.

[ANaaim]

That would be to do something like that for doing inverse dynamics :

    cut_freq = 6

    w = cut_freq / (freq / 2)  # Normalize the frequency
    b, a = signal.butter(4, w, 'low')

    Qdopt_sqp = np.gradient(Qopt_sqp,1/freq, axis=1)
    Qdopt_sqp_filt = signal.filtfilt(b, a, Qdopt_sqp, axis=1)
    Qddopt_sqp = np.gradient(Qdopt_sqp_filt,1/freq, axis=1)
    Qddopt_sqp_filt = signal.filtfilt(b, a, Qddopt_sqp, axis=1)

    for ind_frame in range(0,Qopt_sqp.shape[1]):
        tuple_of_Q = [
            SegmentNaturalCoordinates.from_components(
                u=Qopt_sqp[i*12:i*12+3,ind_frame], rp=Qopt_sqp[i*12+3:i*12+6, ind_frame], rd=Qopt_sqp[i*12+6:i*12+9, ind_frame], w=Qopt_sqp[i*12+9:i*12+12, ind_frame]
            )
            for i in range(0, model.nb_segments)
        ]

        tuple_of_Qddot = [
            SegmentNaturalAccelerations.from_components(
                uddot=Qddopt_sqp_filt[i * 12:i * 12 + 3, ind_frame], rpddot=Qddopt_sqp_filt[i * 12 + 3:i * 12 + 6, ind_frame],
                rdddot=Qddopt_sqp_filt[i * 12 + 6:i * 12 + 9, ind_frame], wddot=Qddopt_sqp_filt[i * 12 + 9:i * 12 + 12, ind_frame]
            )
            for i in range(0, model.nb_segments)
        ]

        Q = NaturalCoordinates.from_qi(tuple(tuple_of_Q))
        Qddot = NaturalAccelerations.from_qddoti(tuple(tuple_of_Qddot))

        torques, forces, lambdas = model.inverse_dynamics(Q=Q, Qddot=Qddot)

[Ipuch]

There is a thin line to determine what should bioNC do.

I was initially more focus on the physical part. So handling the equation Of motion.

My point is that embedding automatic filtering options, I am not convinced yet.

However we could have differentiation module in coordinates class such as:

NaturalCoordinate.differentiate(method="central",dt=...)

And return NaturalVelocity object. :)

As there was two features presented in your post let's start with this one first.

We will see for the other one later.

Also if your priority is to make an example of ID. You can choose to push it entirely and choose to treat this issue later on. If ID example is in the example folder I have no problem to push a code that include custom functionalities.

[ANaaim]

Currently, i would rather focus to the inverse dynamics (in a way i have to go deeper in the code and it is good to do). Let's keep this for later :)

But at least, it was normal to not found such function :)

[Ipuch]

Great !

Don't hesistate to open your pull request.

We'll need to open an issue (a good first) based on our conversation too 👌🏼