Why are there different values between data.M and data.Ycrb ?

[hxrkpib]

Hello,
I have a confuse when calculating the inertia. I load model from a urdf file, then run bellow code:

pinocchio::crba(*model_floating_, *data_floating_, q);
data_floating_->M.triangularView<Eigen::StrictlyLower>() = data_floating_->M.transpose().triangularView<Eigen::StrictlyLower>();
std::cout << "M" << std::endl << data_floating_->M.topLeftCorner(6, 6) << std::endl;
std::cout << "ycrb1" << std::endl << data_floating_->Ycrb[1] << std::endl;

I get common mass value, but there is inertia value bias between data.M and data.Ycrb.

[jcarpent]

The inertia quantity in the spatial inertia is expressed around the center of mass, while it is expressed in the local joint frame for the mass matrix.