opposite results from compute group/group

Dear Lammps users,

Here I have a question about compute group/group. I used compute group/group to get the force between two groups A and B. I tried the following two ways:
compute F1 A group/group B
compute F2 B group/group A
I found that there were opposite results (they have the same absolute value, however one is positive, the other is negative) in these two ways. I do not know if anyone else has the same problem. I wonder if there is special rule that we should follow which group should be in front of the other in this command.
Thank you so much

Kaimin

Dear Lammps users,

Here I have a question about compute group/group. I used compute group/group to get the force between two groups A and B. I tried the following two ways:
compute F1 A group/group B
compute F2 B group/group A
I found that there were opposite results (they have the same absolute value, however one is positive, the other is negative) in these two ways. I do not know if anyone else has the same problem. I wonder if there is

This is not a problem, but the way it has to be. This is Newton's third law.

special rule that we should follow which group should be in front of the other in this command.

The force has to be the force on the first group due to the second.
This is clearly stated in the documentation.

Axel