Question regarding L2 (1s) and (2s) calculations

[Fhrozen]

Hello there.

Thank you for your great work.

I would like to ask about the calculation of ade for 1s, 2s, and 3s. In the case of 1s, the ade is calculated from the index 1 (instead of 0) of pred_traj[1:pred1_len+1], while for 2s and 3s is from index 0 pred_traj[:pred2_len].
Is this a particular situation for calculating L2 on 1s, or maybe a miss in the ade formulation.
Sharing the reference is appreciated.
Best

OpenEMMA/main.py

Lines 461 to 471 in 8403ea6

	pred1_len = min(pred_len, 2)
	ade1s = np.mean(np.linalg.norm(fut_ego_traj_world[:pred1_len] - pred_traj[1:pred1_len+1] , axis=1))
	ade1s_list.append(ade1s)
	pred2_len = min(pred_len, 4)
	ade2s = np.mean(np.linalg.norm(fut_ego_traj_world[:pred2_len] - pred_traj[:pred2_len] , axis=1))
	ade2s_list.append(ade2s)
	pred3_len = min(pred_len, 6)
	ade3s = np.mean(np.linalg.norm(fut_ego_traj_world[:pred3_len] - pred_traj[:pred3_len] , axis=1))
	ade3s_list.append(ade3s)