The ABINIT Users Mailing List ( CLOSED )

["Jin Zhang" <jin.zhang.pku@gmail.com>]

Hi all,

In a response calculation, I noticed the number of bands readin from disk file _WFQ by routine newkpt reported that on some kpoint (0.0 0.0 1/4) only possess half number of bands. There's noting wrong with the result. However, I want to know why is this to fulfill my curiosity. The relevant parts of input and output are listed below... 

&INP--response part--wfk and wfq are obtain from dataset 1 and 2
rfatpol3  1 1
rfdir3  1 0 0
rfphon3  1
tolvrs3   1.0e-12
getwfk3  1
getwfq3  2
kptopt3  3
nqpt3  1
qpt3   0.0 0.0 0.0

nband 10
ngkpt 1 1 16

&OUT---note the kpt  5 has only read 5 bands while others have 10
           ---this kpt corresponds to 0.0 0.0 0.25
 loper3 : total number of electrons, from k and k+q
  fully or partially occupied states are    8.000000E+00 and    8.000000E+00.
 Initialisation of the first-order wave-functions :
  ireadwf=   0
P newkpt: treating     10 bands with npw=   19197 for ikpt=   1 by node    0
P newkpt: treating     10 bands with npw=   19181 for ikpt=   2 by node    0
P newkpt: treating     10 bands with npw=   19121 for ikpt=   3 by node    0
P newkpt: treating     10 bands with npw=   19129 for ikpt=   4 by node    0
P newkpt: treating      5 bands with npw=   19081 for ikpt=   5 by node    1
P newkpt: treating     10 bands with npw=   19077 for ikpt=   6 by node    1
P newkpt: treating     10 bands with npw=   19069 for ikpt=   7 by node    1
P newkpt: treating     10 bands with npw=   19024 for ikpt=   8 by node    1
P newkpt: treating     10 bands with npw=   18980 for ikpt=   9 by node    1

best
Jin Zhang
-- 
Dept. of Physics, Peking University
Beijing, P.R.China
Phone: 86-10-62753190