Compute stress tally

Hi all,
I am using lammps for MD simulations, and when I want to compute stress/tally, lammps shows me an error that " compute stress/tally used with incompatible pair stayle" and my pair style is eam. I was wondering if you could please help me to find an approach to solve this problem.

Thanks,
Elham

#set dimension and boundry condition
dimension 3
boundary p p p

#set units type,atom type and neigborhood
units metal
atom_style atomic
##neigh_modify every 1

#create big box
lattice fcc 3.52
region big sphere 0.0 0.0 0.0 100.0 units lattice
create_box 1 big

#create main atoms
region eli sphere 0.0 0.0 0.0 25.0 side in
create_atoms 1 region eli

group e1 region eli

#create second atoms
region hap sphere 50.5 0.0 0.0 25.0 side in
create_atoms 1 region hap

group e2 region hap

#create third atoms
region cap sphere 25.0 50.0 0.0 25.0 side in
create_atoms 1 region cap

group e3 region cap

#add all the nanoparticles

group all union e1 e2 e3

#set the mass of atoms
mass 1 58.69
##mass 2 58.69
##mass 3 58.69

#set the velocity of atoms
velocity all create 1000 4928459 rot yes mom yes

#define the eam
pair_style eam
pair_coeff * * Ni_u3.eam
##pair_coeff 1 1 0.7 2.517 ##For LJ

#Specify build neighborlist, use cutoff+0.3 A
neighbor 2 bin

#Frequency to rebuild neighborlist
neigh_modify every 1 delay 0 check yes

fix pip all nvt temp 1000 1000 0.005

minimize 1.0e-6 1.0e-8 1000 1000

#set timestep
timestep 0.0005 #0.5 femtosecond

compute nan all stress/tally all

#specify values to show in run
thermo 1000
thermo_style custom step temp pe ke etotal

dump Nidump all atom 1000 nano-3sphere-temp1400.dump
dump Ni all local c_nan 1000 nano1-sphere-temp1400.stress
###dump 1 all custom 1 sphere11_temp1000.txt id type x y z

#specify the method of calculation

run 60000

Hi all,
         I am using lammps for MD simulations, and when I want to compute
stress/tally, lammps shows me an error that " compute stress/tally used with
incompatible pair stayle" and my pair style is eam. I was wondering if you
could please help me to find an approach to solve this problem.

there is not much of a point of using compute stress/tally on group
all with group all, and the documentation already warns against using
stress/tally with manybody potentials.

just use compute stress/atom.

please also note, that stress/tally does not compute a local property.

axel.