Data approximation strategies between generalized line scales and the influence of labels and spacing