Re: [Wien] regarding 2Doptimization
20.03.2017 23:38, Dr. K. C. Bhamu wrote:
vol_-15.0, vol_-10.0, vol_-5.0, vol_-0.0, vol_5.0
and few scf steps for vol_10.
Now I wanted to run the case from the last point from where it was
stopped (from vol_5).
I know that if I use -NI switch at the end of the runsp_lapw .
script then it should start from the point where it was stopped.
Your understanding is not quite correct: you have an scf procedure with
many cycles, after each cycle some history being kept. The -NI means
that program (run_lapw) does not remove this history (files with .broyd
and some others).
But, in my case it starting from vol_-10 instead of vol_5.
My opimize.job file is below:
... foreach i ( \
case_vol_-15.0 \
> ... cp ${i}.struct case.struct
this means than program should take first case_vol_-15.0 . If you want
vol_5 edit this place.
# cp ${i}_default.clmsum case.clmsum >>> I aslo tried to uncomment
the these lines (once for non FM and then in another test for FM case)
but nothing has change.
You should do this command in order to start with the density of the
case that you wish.
clmextrapol_lapw
this makes extrapolation of the density from your last struct file
(judging from your words the last was vol_10) to the new one (in your
case vol_-15.0)
min -I -j 'runsp_lapw -p -ec 0.0001 -fc 2 -i 100 -NI'
-NI here is a strong error if occasionally the .broyd files were not
deleted: you have a new case but use a history from another case.
Best wishes
Lyudmila Dobysheva
--
Phys.-Techn. Institute of Ural Br. of Russian Ac. of Sci.
426001 Izhevsk, ul.Kirova 132
RUSSIA
--
Tel.:7(3412) 432045(office), 722529(Fax)
E-mail: l...@ftiudm.ru, lyuk...@mail.ru (office)
lyuk...@gmail.com (home)
Skype: lyuka17 (home), lyuka18 (office)
http://ftiudm.ru/content/view/25/103/lang,english/
--
___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
[Wien] regarding 2Doptimization
Dear Wien2k users,
I am optimizing a case for FM structure on 32 cores (two nodes). My job was
hangged after optimizing the structure for
vol_-15.0, vol_-10.0, vol_-5.0, vol_-0.0, vol_5.0
and few scf steps for vol_10.
Now I wanted to run the case from the last point from where it was stopped
(from vol_5).
I know that if I use -NI switch at the end of the runsp_lapw . script
then it should start from the point where it was stopped.
But, in my case it starting from vol_-10 instead of vol_5.
Please suggest me what mistake I am doing.
My opimize.job file is below:
#!/bin/csh -f
# Modify this script according to your needs:
# switches in run_lapw or runsp_lapw,
# spin-polarization (change run_lapw to runsp_lapw)
# modify the save_lapwcommand
if (-e case.clmsum && ! -z case.clmsum) then
x dstart -super -p
endif
if (-e case.clmup && ! -z case.clmup) then
x dstart -super -up -p
x dstart -super -dn -p
endif
foreach i ( \
case_vol_-15.0 \
case_vol_-10.0 \
case_vol__-5.0 \
case_vol___0.0 \
case_vol___5.0 \
case_vol__10.0 \
)
rm case.struct # NFS-bug
cp ${i}.struct case.struct
# Please uncomment and adapt any of the lines below according to your
needs
# if you have a previous optimize-run:
# cp ${i}_default.clmsum case.clmsum >>> I aslo tried to uncomment the
these lines (once for non FM and then in another test for FM case) but
nothing has change.
# cp ${i}_default.clmup case.clmup
# cp ${i}_default.clmdn case.clmdn
# if you want to start with dstart:
#x dstart -p
#x dstart -up -p
#x dstart -dn -p
# recommended default method: use charge extrapolation
clmextrapol_lapw
if (-e case.clmup && ! -z case.clmup) then
clmextrapol_lapw -up
clmextrapol_lapw -dn
endif
# modify the run_lapw command below
# (parallel, convergence, iter.diag, MSR1a minimization,...) or
# comment the run_lapw line and activate spinpolarization
#runsp_lapw -ec 0.0001
#min -I -j "run_lapw -I -fc 1.0 -i 40 "
min -I -j 'runsp_lapw -p -ec 0.0001 -fc 2 -i 100 -NI'
set stat = $status
if ($stat) then
echo "ERROR status in" $i
exit 1
endif
# Typically one would change below: default --> pbe_1000k_rkm7
save_lapw -f ${i}_default
#save_lapw -f -d XXX ${i}
end
Regards
Bhamu
___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
