The Point struct stores redundant information

[ufechner7]

The Point struct is defined as:

struct Point
    idx::Int16
    transform_idx::Int16 # idx of wing used for initial orientation
    wing_idx::Int16
    pos_cad::KVec3
    pos_b::KVec3 # pos relative to wing COM in body frame
    pos_w::KVec3 # pos in world frame
    vel_w::KVec3 # vel in world frame
    type::DynamicsType
end

It stores three position vectors, even though a point can only have one and only one position.

This smells.

[1-Bart-1]

It is not redundant.
Well, pos_b is used if it is a WING point. And pos_w and pos_cad are used otherwise.

[ufechner7]

Please, explain. It is just mathematical logic. A point can only have one position. And if you need the position in different reference frames, this should be done on-the-fly. Otherwise there is the risk of inconsistencies.

Data Inconsistency: When the same piece of data exists in multiple places, updates or changes may not be applied uniformly, leading to conflicting or outdated information within the system. This inconsistency undermines the reliability and accuracy of your data.

[1-Bart-1]

Just one if these positions is being used at each time, so there is no issue. pos_b when the point is of type WING, pos_w to initialise the mtk simulation and it can be updated each time step with current pos from simulation. pos_cad is used to describe the system before transform is applied, and it is not used in simulation.

[1-Bart-1]

pos_cad is constant, pos_b and pos_w are updated during sim