Well the LJ potential table that I generated seems totally fine. The interaction values are all positive below the sigma value and the they become negative as the r increases aboce the sigma. However the potential table that I obtained using the pair write command from the simulation in which I used the LJ potential directly, doesn't look anything like this. the only positive value is the first value and then they all become negative. I have attached a both potential tables the one I generated and the one which lammps generated for lj. It seems as if lammps is doing something extra to the potential table besides only computing the interactions using the LJ formula, to make it applicable to a lammps simulation.
table.potential.lammps.txt (581 KB)
table.potential.mytable.txt (355 KB)