A possible typo in a doc formula of chainjnttojacdotsolver

[torydebra]

Hi, I was just wondering about some documentation in the chainjnttojacdotsolver

orocos_kinematics_dynamics/orocos_kdl/src/chainjnttojacdotsolver.hpp

Lines 153 to 165 in 88d5c8f

	/**
	* @brief Computes \f$ \frac{\partial J^{i,ee}}{\partial q^{j}}.\dot{q}^{j} \f$
	*
	* @param bs_J_bs The Jacobian expressed in the base frame with the end effector as the reference point
	* @param joint_idx The indice of the current joint (j in the formula)
	* @param column_idx The indice of the current column (i in the formula)
	* @param representation The representation (Hybrid,Body-fixed,Inertial) in which you want to get dJ/dqj .qdotj
	* @return Twist The twist representing dJi/dqj .qdotj
	*/
	const Twist& getPartialDerivative(const Jacobian& J,
	const unsigned int& joint_idx,
	const unsigned int& column_idx,
	const int& representation);

I see that the formula written is $ \frac{\partial J^{i,ee}}{\partial q^{j}}.\dot{q}^{j} $, but I am not sure that the final $\dot{q}^{j}$ should be there.
Indeed in :

orocos_kinematics_dynamics/orocos_kdl/src/chainjnttojacdotsolver.cpp

Lines 113 to 115 in 88d5c8f

	// Column J is the sum of all partial derivatives ref (41)
	if(!locked_joints_[j])
	jac_dot_k_ += getPartialDerivative(jac_,j,k,representation_) * q_in.qdot(j);

The result of getPartialDerivative is multiplied by qdot(j)

I am using these formulas/codes, so I would just like to be sure about what it is happening inside :D

Thanks!

[MatthijsBurgh]

I don't have the formulas ready to me. @meyerj @smits could you take this one?

[traversaro]

I guess also @ahoarau may know something as the original author of the class.