[Wien] [by mistake I replies in another thread] problem in 2Doptimize.job or in my method ...

Fecher, Gerhard fecher at uni-mainz.de
Thu Jan 12 08:59:21 CET 2017


It seems you still do not know what the nn error is telling you.

How much is the nearest neighbour distance reduced when you reduce from the initial structure the volume by 15 % and afterwards c/a by 9%
(My guess would be more than 10%)
(in addition what happens if you shift the atom with the free parameter)

Try to initialize the calculation for the structure that causes the nn error and see what happens.
(also try to initialize the structure with the largest volume and c/a reduction.)

Which atom is affected when you use min ?
How does this influence the nearest neighbour distances (which ones) ?

What do you finally conclude for your RMT settings ?

Ciao
Gerhard

DEEP THOUGHT in D. Adams; Hitchhikers Guide to the Galaxy:
"I think the problem, to be quite honest with you,
is that you have never actually known what the question is."

====================================
Dr. Gerhard H. Fecher
Institut of Inorganic and Analytical Chemistry
Johannes Gutenberg - University
55099 Mainz
and
Max Planck Institute for Chemical Physics of Solids
01187 Dresden
________________________________________
Von: Wien [wien-bounces at zeus.theochem.tuwien.ac.at] im Auftrag von Dr. K. C. Bhamu [kcbhamu85 at gmail.com]
Gesendet: Mittwoch, 11. Januar 2017 22:15
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] [by mistake I replies in another thread] problem in 2Doptimize.job or in my method ...

Thank you Prof. Gerhard and Lyudmila
Please see my additional information (some additional queries are there):


what switches do you usually use when you run a LDA+U calculation ?
  min  -j "runsp_c_lapw -orb -I -fc 1.0 -i 90 "

why don't you use them here if you like to do a LDA+U calculation ?
Now, I corrected it.

you have a NN error, when does this error occur in general ?

It occurs at the beginning for
ERROR status in 2D_V-15.0_COA_-9.0   (-15 -10 -5 0 5 10 15 Vs. -9 -6 -3 0 3 6 9)
ERROR status in 2D_V-10.0_COA_-9.0 (-10 -5 0 5 10 vs. -9 -6 3 0 3 6 9)
ERROR status in 2D_V-10.0_COA_-8.0 (-10 -5 0 5 10 Vs. -8 -6 -4 -2 0 2 4 6 8)
ERROR status in 2D_V-10.0_COA_-6.0 (-10 -5 0 5 10 Vs.  -6 -4 -2 0 2 4 6 )
ERROR status in 2D_V-8.0_COA_-6.0 (-8 -6 -4 -2 0 2 4 6 8 Vs. -6 -4 -2 0 2 4 6)
ERROR status in 2D_V-6.0_COA_-6.0 (-6 -4 -2 0 2 4 6 Vs. -6 -4 -2 0 2 4 6)

The simplest  (with less parameters) case was -6 -4 -2 0 2 4 6" for vol and  -6%,-3%,0,3%,6% (also -6 -4 -2 0 2 4 6 ) for c/a  where I encountered the nn error.

Now, as per suggestions of Dr. Lyudmila, I tried to see in which file the problem was. So:

In the new test, I did not encounter the nn problem for "-6 -4 -2 0 2 4 6" volume changes and -4, -2, 0 2 4 for c/a.
Should I take this small value only?  I wanted to optimise it for some more parameters but it giving nn error beyod these range.


how much did you reduce the rmt's and what is the nearest neighbour distance for the structure with smallest volume and c/a ratio ?

 I reduced 5% rmts and the attached one is with this reduced rmt (CuGaO2.struct) and nn distance is;

ATOM  1  Cu         ATOM  3  O
 RMT(  1)=1.75000 AND RMT(  3)=1.50000
 SUMS TO 3.25000  LT.  NN-DIST= 3.43911

    ATOM  2  Ga         ATOM  3  O
 RMT(  2)=1.85000 AND RMT(  3)=1.50000
 SUMS TO 3.35000  LT.  NN-DIST= 3.73567

    ATOM  3  O          ATOM  1  Cu
 RMT(  3)=1.50000 AND RMT(  1)=1.75000
 SUMS TO 3.25000  LT.  NN-DIST= 3.43911

For original struct, the nn distance is below (CuGaO2_orig.struct):

    ATOM  1  Cu         ATOM  3  O
 RMT(  1)=1.84000 AND RMT(  3)=1.58000
 SUMS TO 3.42000  LT.  NN-DIST= 3.43911

    ATOM  2  Ga         ATOM  3  O
 RMT(  2)=1.95000 AND RMT(  3)=1.58000
 SUMS TO 3.53000  LT.  NN-DIST= 3.73567

    ATOM  3  O          ATOM  1  Cu
 RMT(  3)=1.58000 AND RMT(  1)=1.84000
 SUMS TO 3.42000  LT.  NN-DIST= 3.43911



what happens if you change the position of one of the atoms ?

How much I change (+/-?) and for which? would you plz suggest, I will apply and test it.


is the struct file the original one or the one where the calculation fails ?

CuGaO2.struct is with 5% rmt reduction and CuGaO2_orig.struct is original.


If you change the script then you should know what you are doing, usually it helps to read the manual to find out what the commands mean
a basic knowledge on linux commands is also helpful, what means cp ?

I am sorry. cp just copy one file to other and I understood now it from Prof. Lyudmila's reply. I should not change it.

Did you intend to run the scf cycles several times for the same structure ? I guess not.

No, I did not.  I want to do 2Doptimisation and then want to apply mBJ potential with the optimised one.

As Ludmilla told, it doesn't make sense to arbitrarily delete commands from the script by commenting them,
in particular, if those commands do not have anything to do with the error you receive.

I realised it now. Thank you for correcting me.


Sincerely
Bhamu
Ciao
Gerhard

DEEP THOUGHT in D. Adams; Hitchhikers Guide to the Galaxy:
"I think the problem, to be quite honest with you,
is that you have never actually known what the question is."

====================================
Dr. Gerhard H. Fecher
Institut of Inorganic and Analytical Chemistry
Johannes Gutenberg - University
55099 Mainz
and
Max Planck Institute for Chemical Physics of Solids
01187 Dresden
________________________________________
Von: Wien [wien-bounces at zeus.theochem.tuwien.ac.at<mailto:wien-bounces at zeus.theochem.tuwien.ac.at>] im Auftrag von Dr. K. C. Bhamu [kcbhamu85 at gmail.com<mailto:kcbhamu85 at gmail.com>]
Gesendet: Dienstag, 10. Januar 2017 18:18
An: A Mailing list for WIEN2k users
Betreff: [Wien] problem in 2Doptimize.job or in my method ...

Dear Users
I am running the attached struct file for 2Doptimization. The structure is hexagonal.
I am running it for PBE+U.

First query:
Is it ok to use it for PBE+U ( min -j "runsp_lapw -p -i 200 -ec 0.00005 -cc 0.0005 -fc 0.5 -I") ??

when I submit the 2Doptimize.job  my initialised struct file changed to the initial one and it gives me "nn error".
then I tried to comment/uncomment many lines in the script (dstart and clmextrapol_lapw) but the problem did not solve. Finally, I  uncommitted the line "## cp $i.struct CuGaO2.struct" see below in blue with bold. Now the job is running well.

Second query:
My query is whether what I did is fine of I should not change anything in the script. If I do not change anything in the script then the nn problem occurs.


 #ana2D_lapw#!/bin/csh -f
 #
 #Modify this script according to your needs
 unalias rm
 #
 # to reuse previous scf runs (without a new scf run) set answscf=y
 # and use the same "savename".
 #
 # When you make modifications (RKmax, k-mesh, XC-potentials) choose:
 # answscf=no, old_savename=savename and a new savename (eg. "_pbe_rk8_1000k").
 set answscf=y
 set savename=
 set old_savename=
 #
 set numbvcoa = `head -1 < numbvcoa`
 set numbcoa = $numbvcoa[2]
 #
 rm -f CuGaO2.Vconst*
 if !(-d Vconst$savename) then
    mkdir Vconst$savename
 endif
 if (-e CuGaO2.clmsum && ! -z CuGaO2.clmsum) then
   x dstart -super
 endif
 if (-e CuGaO2.clmup && ! -z CuGaO2.clmup  ) then
   x dstart -super -up
   x dstart -super -dn
 endif

 set count = 1
 set j = 0
  @ j = $j + 1
 set jj = 1
 foreach i ( \


                                       2D_V__-15.0_COA__9.0 \
                                      *****************
                                       *****************

                                       2D_V_15.0_COA__9.0 \
   )
 echo "*******************************"
 echo $i
 echo "*******************************"
 if ( -e "${i}$savename.scf" ) then
   if ($answscf == "y" ) then
      echo "Reusing old ${i}$savename.scf file."
   else
      echo "Old ${i}$savename.scf file will not be used."
   endif
   if ($answscf == "y" )  goto exitscf
 endif
## cp $i.struct CuGaO2.struct                                 changed from  cp $i.struct CuGaO2.struct     to    ## cp $i.struct CuGaO2.struct

 # if you have a previous optimize-run:
 #    cp  ${i}$old_savename.struct CuGaO2.struct
 #    cp  ${i}$old_savename.clmsum CuGaO2.clmsum
 #    cp  ${i}$old_savename.clmup CuGaO2.clmup
 #    cp  ${i}$old_savename.clmdn CuGaO2.clmdn
 # if you want to start with dstart:
 #    x dstart   # -c
 #    x dstart -up  # -c
 #    x dstart -dn  # -c
 # recommended option: use charge extrapolation
 clmextrapol_lapw
 if (-e CuGaO2.clmup  && \
  ! -z CuGaO2.clmup  ) then
     clmextrapol_lapw -up
     clmextrapol_lapw -dn
 endif

 #I HIGHLY RECOMMEND TO RELAX YOUR STRUCT FILE AND THEN USE IT.

 #Usually low-symmetry compounds have internal coordinates
 # and you must relax them (activate -min).

 #run_lapw -ec 0.0001   # -p -it -min -fc 1.0 -cc 0.01
  min -j "runsp_lapw -p -i 200 -ec 0.00005 -cc 0.0005 -fc 0.5 -I"
 # runsp_lapw -ec 0.0001

     set stat = $status
     if ($stat) then
        echo "ERROR status in" $i
        exit 1
     endif

 save_lapw  -f ${i}$savename

 exitscf:
 grepline :ENE "${i}$savename.scf" 1 > analysisENE
 grepline :VOL "${i}$savename.scf" 1 > analysisVOL
 set ene=`grep :ENE analysisENE | cut -f2 -d= `
 set vol=`grep :VOL analysisVOL | cut -f2 -d= `
 set a=`head -4 < ${i}$savename.struct | tail -1|cut -c1-9`
 set c=`head -4 < ${i}$savename.struct | tail -1|cut -c22-30`
 if ( $count > $numbcoa ) then
 @ j = $j + 1
 set numbvcoa = `head -$j < numbvcoa | tail -1`
 set numbcoa = $numbvcoa[2]
 set count = 1
 endif
 set VCOA=`echo "$i" | cut -c4-18`
 set coa=`echo " $c / $a " | bc -l`
 echo $c $a $vol $ene $coa  $VCOA>>CuGaO2.Vconst$j
 @ count = $count + 1
 @ jj = $jj + 1

 end
 cp CuGaO2.Vconst* Vconst$savename
 echo " all CuGaO2.Vconst* files are also saved in Vconst$savename "
 echo " use  ana2D_lapw   to analyze the results"
 #ana2D_lapw

Sincerely
Bhamu
_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at<mailto:Wien at 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



More information about the Wien mailing list