problem with calculating viscousity

dear all
I am new in lammps and i try to calculate the viscousity of water molecular (spce) confined between two harmonic walls. in lammps manual, a Sample LAMMPS input script has been written for viscosity of liquid Ar. so i tried to modify this code for my simulation. so my lammps input script is:

variable T equal 300
variable V equal vol
variable dt equal 1
variable p equal 400
variable s equal 5
variable d equal $p*$s

variable kB equal 1.3806504e-23 # [J/K] Boltzmann

variable A2m equal 1.0e-10
variable fs2s equal 1.0e-15
variable atm2Pa equal 101325.0
variable convert2 equal {atm2Pa}*{atm2Pa}{fs2s}*{A2m}{A2m}*{A2m}
dimension 3
include system.in.init

read_data system.data

include system.in.settings
velocity all create $T 34387 rot yes dist gaussian

fix 55 up-wall nvt temp $T $T 10
fix 5 down-wall nvt temp $T $T 10
fix 4 spce nve

fix 7 spce temp/rescale 100 $T $T 0.02 1.0
fix 77 up-wall temp/rescale 100 $T $T 0.02 1.0
fix 777 down-wall temp/rescale 100 $T T 0.02 1.0 timestep {dt}

thermo $d
thermo_style custom step temp pe ke etotal enthalpy press vol ebond ecoul eangle
run 20000

variable pxy equal pxy

variable pxz equal pxz

variable pyz equal pyz

fix SS spce ave/correlate $s $p $d v_pxy v_pxz v_pyz type auto file S0St.dat ave running

variable scale2 equal {convert2}/({kB}$T)$V*s*{dt}

variable v11 equal trap(f_SS[3])*${scale2}

variable v22 equal trap(f_SS[4])*${scale2}

variable v33 equal trap(f_SS[5])*${scale2}

thermo_style custom step temp press v_pxy v_pxz v_pyz v_v11 v_v22 v_v33
run 10000

variable v equal (v_v11+v_v22+v_v33)/3.0
variable ndens equal count(spce)/vol

variable k equal (v_k11+v_k22+v_k33)/3.0

variable ndens equal count(spce)/vol

print “average viscosity: $v [Pa.s/ @ T K, {ndens2} /A^3”

my problem is pxy,pxz and pyz. because i think these values are for all atoms (both wall and water molecules) and i just want to calculate the viscousity of water molecule (no walls).
how can i calculate the pxy pxz and pyz just for water (no water+wall)??

can i use compute stress/atom command instead?? if yes how should i modify the fix ave/correlate???
can i use the flowing code??
:
:

compute mystress spce stress/atom

fix SS spce ave/correlate $s $p $d c_mysterss[4] c_mysterss[5] c_mysterss[6] type auto file S0St.dat ave running
:
: