May compute temperture have little problem?

Hi everyone,
At first, I apologize to Axel and Steve for my rudeness.

I extract my problem, I bulid Kremer-Grest model, then name the end of chain type 2 while the other type1

Then I calulcate the temperture of all atoms(this command is predefined,there are 13200 particles) and of type1(there are 12144 particles)

I think the temperture should be same in equlibrium

But the log file give me the answer is not

Here it is,

LAMMPS (7 Dec 2015)
#=================Some variable===================
variable Rcut equal 2^(1.0/6.0) #Rcut in LJ potential
variable deltaT equal 0.005 #step length

variable Nstep equal 100 #How many steps to print Rg et.al
variable Noutput equal 1000 #How many steps to output coordinate
variable Ntotal equal 1000 #total timestep

#================Basic information================
units lj
atom_style bond
special_bonds fene
comm_modify vel yes #This is important for ghost particles in DPD

#===============Set initial configuration==========
read_data data.chain
orthogonal box = (-12.4745 -12.4745 -12.4745) to (12.4745 12.4745 12.4745)
1 by 1 by 2 MPI processor grid
reading atoms …
13200 atoms
reading velocities …
13200 velocities
scanning bonds …
1 = max bonds/atom
reading bonds …
12672 bonds
2 = max # of 1-2 neighbors
2 = max # of special neighbors

#=====================Force field==================
pair_style hybrid/overlay lj/cut {Rcut} dpd/tstat 1.0 1.0 {Rcut} 22046
pair_style hybrid/overlay lj/cut 1.12246204830937 dpd/tstat 1.0 1.0 ${Rcut} 22046
pair_style hybrid/overlay lj/cut 1.12246204830937 dpd/tstat 1.0 1.0 1.12246204830937 22046
pair_coeff * * lj/cut 1.0 1.0
pair_modify shift yes
pair_coeff * * dpd/tstat 0.5

bond_style fene
bond_coeff * 30.0 1.5 1.0 1.0

group molFree type 1
12144 atoms in group molFree
group molEnd type 2 #The end of polymer is type 2(every chain has two end monomer)
1056 atoms in group molEnd
#===================Neighbourlist===================
neighbor 0.4 bin
neigh_modify delay 0 every 1 check yes

#====================Output=========================
compute molChunk all chunk/atom molecule
compute molCen all com/chunk molChunk #calculate chain center of each Chunk
compute TTTT molFree temp #calculate temperture about “free” monomers

fix 1 all nve
fix 2 all ave/time 1 1 ${Nstep} c_molCen file “CAL_CHAIN CENTER.DAT” mode vector format “%21.14f”
fix 2 all ave/time 1 1 100 c_molCen file “CAL_CHAIN CENTER.DAT” mode vector format “%21.14f”

dump 1 molEnd custom ${Nstep} CAL_END.DAT id xu yu zu
dump 1 molEnd custom 100 CAL_END.DAT id xu yu zu
dump_modify 1 sort id format “%7d %21.14f %21.14f %21.14f”

dump 2 all custom ${Noutput} POSITION_.DAT id x y z
dump 2 all custom 1000 POSITION_
.DAT id x y z
dump_modify 2 sort id format “%7d %21.14f %21.14f %21.14f”

dump 3 all custom ${Noutput} RE-POSITION_.DAT id xu yu zu
dump 3 all custom 1000 RE-POSITION_
.DAT id xu yu zu
dump_modify 3 sort id format “%7d %21.14f %21.14f %21.14f”

dump 4 all custom ${Noutput} VELOCITY_.DAT id vx vy vz
dump 4 all custom 1000 VELOCITY_
.DAT id vx vy vz
dump_modify 4 sort id format “%7d %21.14f %21.14f %21.14f”

thermo ${Nstep}
thermo 100
thermo_style custom step temp press c_TTTT[1] c_TTTT[2] c_TTTT[3]

#====================Other set=========================

timestep ${deltaT}
timestep 0.005

run ${Ntotal}
run 1000
Neighbor list info …
2 neighbor list requests
update every 1 steps, delay 0 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 1.52246
ghost atom cutoff = 1.52246
binsize = 0.761231 -> bins = 33 33 33
Memory usage per processor = 10.9025 Mbytes
Step Temp Press TTTT[1] TTTT[2] TTTT[3]
0 1.0097502 4.9765318 0.91726008 0.94104676 0.94831042
100 1.0018254 5.1593168 0.90467855 0.91600982 0.93486771
200 1.0062245 5.0518647 0.91199219 0.94559787 0.91859349
300 1.0089064 4.8540648 0.91904713 0.92769432 0.94068556
400 1.0050043 4.8619678 0.91900574 0.91592041 0.93482699
500 1.0089963 5.0404991 0.92568045 0.9231669 0.93217811
600 1.0012736 5.1192684 0.92122697 0.91607784 0.93397356
700 1.0022566 5.1291152 0.9105485 0.93217761 0.92706954
800 1.0067518 4.9855736 0.9280562 0.93180069 0.91991404
900 1.0033721 5.0171879 0.90998295 0.91147157 0.94753534
1000 1.0083134 4.9329235 0.93777529 0.93510241 0.90926221
Loop time of 3.78962 on 2 procs for 1000 steps with 13200 atoms

Hi everyone,
    At first, I apologize to Axel and Steve for my rudeness.

    I extract my problem, I bulid Kremer-Grest model, then name the end of
chain type 2 while the other type1

    Then I calulcate the temperture of all atoms(this command is
predefined,there are 13200 particles) and of type1(there are 12144
particles)

    I think the temperture should be same in equlibrium

it *is*.

    But the log file give me the answer is not

because you don't seem to output the temperature value of your compute
anywhere. in your thermo_style line, you refer not to the global
scalar c_TTTT, but to three of the six terms of the kinetic energy
tensor, c_TTTT[1], c_TTTT[2] and c_TTTT[3].

axel.

   Here it is,

    LAMMPS (7 Dec 2015)

[...]

group molFree type 1
12144 atoms in group molFree
group molEnd type 2 #The end of polymer is type 2(every chain has two end
monomer)
1056 atoms in group molEnd

[...]

compute TTTT molFree temp #calculate temperture about "free" monomers

[...]