[Octopus-users] XCrysden output format

[Robertson Burgess]

Dear Octopus Users,

I've been outputting the electron density data in the XCrysden format. For the first structure it tried it on it worked fine, and the last part of the file before all the data was as follows;

BEGIN_BLOCK_DATAGRID3D
units: coords = A, function = A^-3
DATAGRID_3D_function
     87     87     87
0.0 0.0 0.0
   17.200000    0.000000    0.000000
    0.000000   17.200000    0.000000
    0.000000    0.000000   17.200000

This was exactly as expected, giving my a data grid spacing of 17.2/(87-1)=0.2 exactly. This is exactly the value of my Spacing input, so this is what I expected.
When I ran the same thing for a larger system, however, I got the following block:

BEGIN_BLOCK_DATAGRID3D
units: coords = A, function = A^-3
DATAGRID_3D_function
    117    117    117
0.0 0.0 0.0
   23.457460    0.000000    0.000000
    0.000000   23.457460    0.000000
    0.000000    0.000000   23.457460

which gives me a grid spacing of ~0.202219. However my Spacing value in my inp file is the same as before, exactly 0.2, but that would mean that the size of my grid should instead be 23.2. Why is it for this larger system that the grid spacing has changed? I would have thought the grid spacing should follow exactly from the input in Octopus? The reason that this is important to me is because my smaller system above is actually a sub-section of my larger system, and if the grid spacings are the same then it would be very straight forward to subtract the electron density of my top system from that of my bottom system. If the densities are not the same, that obviously makes things substantially more problematic.

Thanks,
Bob Burgess
University of Newcastle, Australia