# Origin in Lammps, cell, coordinate

Dear All,

A little non-technique question when I try doing some coordinate transformation.

Lammps uses (xlo, xhi, ylo, yhi, zlo, zhi, xy, xz, yz) to define a cell and (xlo, ylo, zlo) is the "origin" of the cell. I was a bit confused about if this "origin" is also the origin that atom coordinates are referred to since the value of this "origin" is not fixed during calculation.

Explicitly, say at some step we have (xlo= -1, ylo=zlo=0) and some atom with coordinates (1,0,0), then the distance between this atom and the "origin" corner of the cell is 1 or 2? I assume it's 2, but the manual didn't explicitly say this (am I right?). This results in my next question: what's the benefit for this floating origin instead of fixing the "origin" at (0,0,0)?

Thanks

Eugene

Dear All,

A little non-technique question when I try doing some coordinate
transformation.

Lammps uses (xlo, xhi, ylo, yhi, zlo, zhi, xy, xz, yz) to define a cell
and (xlo, ylo, zlo) is the "origin" of the cell. I was a bit confused
about if this "origin" is also the origin that atom coordinates are
referred to since the value of this "origin" is not fixed during
calculation.

Explicitly, say at some step we have (xlo= -1, ylo=zlo=0) and some atom
with coordinates (1,0,0), then the distance between this atom and the
"origin" corner of the cell is 1 or 2? I assume it's 2, but the manual
didn't explicitly say this (am I right?). This results in my next
question: what's the benefit for this floating origin instead of fixing
the "origin" at (0,0,0)?

Yes, it's 2 units away from the edge of the box at that point. The reason for allowing this is largely one of convenience; there's no good reason why it has to be fixed, so why force the user to make one corner (0,0,0)?

--AEI