Cluster Expansion for Coupled sublattice systems

Alloy Theoretic Automated Toolkit (ATAT)
1

I want to use ATAT with my VASP for doped system with vacancy, namely coupled sublattice system.
Can anyone tell me is that kind of expansion is possible with multicomponents mmaps code ? or there any other options ?
Thanks in advance

2

Thank you ajjackson.
I am using mmaps. I make lat.in file for CexLa(1-x)O(2-y) system. but mmaps warning that I am not using primitive cell, though the code is running.
My lat.in file is as below:

5.51 5.51 5.51 90 90 90
1 0.00 0.00
0.00 1 0.00
0.00 0.00 1
0.0000000000000000 0.0000000000000000 0.0000000000000000 Ce, La
0.5000000000000000 0.5000000000000000 0.0000000000000000 Ce, La
0.5000000000000000 0.0000000000000000 0.5000000000000000 Ce, La
0.0000000000000000 0.5000000000000000 0.5000000000000000 Ce, La
0.2500000000000000 0.2500000000000000 0.2500000000000000 O, Vac
0.2500000000000000 0.7500000000000000 0.2500000000000000 O, Vac
0.7500000000000000 0.2500000000000000 0.7500000000000000 O, Vac
0.7500000000000000 0.7500000000000000 0.7500000000000000 O, Vac
0.7500000000000000 0.2500000000000000 0.2500000000000000 O, Vac
0.2500000000000000 0.7500000000000000 0.7500000000000000 O, Vac
0.2500000000000000 0.2500000000000000 0.7500000000000000 O, Vac
0.7500000000000000 0.7500000000000000 0.2500000000000000 O, Vac

Is it ok to work with this, or should I use a primitive cell for this structure?
Thanks all in advance

3

MMAPS allows you to specify "Vac" as a species - i.e. neutral vacancies. This is included in the man page example https://www.brown.edu/Departments/Engin … ode49.html

4

To find the primitive cell, do

cellcvrt -s < lat.in

For (m)maps, not using the primitive cell could cause the code to skip some structures and make it less efficient (slower).
For the other codes (e.g. (m)emc2), there is no problem in using a nonprimitive lat.in, although efficiency may suffer.