I have two Geodataframes: first one (let\'s call it paths) contains shortest paths (by length) from 16 origin points to 20 destination points, each cell corresponds to one n