Calculation of the get_matrix() method

[adq-24]

Can someone please explain how the transformation matrix is calculated in the transform.get_matrix() method, i.e., in which order the transformations are applied? I haven't found anything online, and when I calculate it myself, I don't get the matrix in the source code.

I am using the following transformation matrices for the transformations in the left-handed coordinate system of CARLA:

• For the rotation around the x-axis:
  
• For the rotation around the y-axis:
  
• For the rotation around the z-axis:
  

In my opinion, the transformation matrix in get_matrix() (

carla/LibCarla/source/carla/geom/Transform.h

Line 103 in a21cc5e

std::array<float, 16> transform = {

) looks like this:

It is unclear to me how or in which order the individual rotation matrices are multiplied to arrive at this matrix. I have found that you get it by calculating T_CARLA = T_z_alt * T_y * T_x, where T_z_alt would be the transformation matrix of a rotation around the z-axis in a right-handed coordinate system instead of the left-handed coordinate system. However, this does not seem logical because CARLA uses a left-handed coordinate system. Could someone please explain what I am missing or where I have made a mistake?