[Wien] Wien2k OPTIC crashing for XMCD calculation

[Gavin Abo]

An additional comment:

I have seen that input conversion error before when IPRINT in case.inc 
was set to 1 for only "some" atoms.  Setting IPRINT = 1 for "all" atoms 
in the case.inc file seemed to remove the error.

On 5/15/2015 1:23 AM, Peter Blaha wrote:
> You cannot do XMCD without spin-orbit coupling.
>
> You NEVER added a line with lapwso in your description of what you 
> actually did and also the lapw2 step (-fermi) is missing.
>
> Please check carefully the UG and previous posts on the EXACT sequence 
> of commands.
>
> If this does not help, send us the EXACT list of commands you entered !!
>
>
> Am 15.05.2015 um 09:01 schrieb Santu Baidya:
>> Dear Prof. Blaha,Gavin Abo andXavier Rocquefelte*,*
>> **  Thank you for your suggestions. As you see from my previous mail I
>> first did GGA+U+SO calculation for XMCD of Co L23 edge.
>>   It did not work and I moved to GGA+U cal.
>>
>>   I first did scf calculation under GGA+U+SO. then ran these commands as
>> i see from forum.....
>> a) runsp -so -dm -orb -c -s lapw1 -e lcore
>> b) x kgen -so  (for dense grid)
>> c) x lapw1 -up/dn -orb -p(for eigenvalues)
>> d) Edit case.inop with xtra line "XMCD 1 L23" as it is first atom
>> e) x optic -so -up
>>
>> Then the problem comes with error.......
>>
>> running OPTIC in parallel mode
>> [1] 24537
>> forrtl: severe (64): input conversion error, unit 35, file
>> /users/santu/ProjectCoxOy/AFM-A-CoO/GGA+U+SO/xmcd/try/CoO/CoO.corewfup
>> Image              PC                Routine Line        Source
>> opticc             00000000004E9ABD  Unknown Unknown  Unknown
>> opticc             00000000004E85C5  Unknown Unknown  Unknown
>> opticc             0000000000492599  Unknown Unknown  Unknown
>> opticc             0000000000451DCA  Unknown Unknown  Unknown
>> opticc             00000000004515C0  Unknown Unknown  Unknown
>> opticc             000000000047126B  Unknown Unknown  Unknown
>> opticc             000000000046E60E  Unknown Unknown  Unknown
>> opticc             000000000043BEFA  cor_mat_ 220
>>   sph-UPcor_tmp.f
>> opticc             000000000041D3A5  MAIN__ 460  opmain.f
>> opticc             00000000004036AC  Unknown Unknown  Unknown
>> libc.so.6          00002AF40916AC36  Unknown Unknown  Unknown
>> opticc             0000000000403589  Unknown Unknown  Unknown
>> [1]  + Done                          ( cd $PWD; $t $exe
>> ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>> [1] 24543
>> OPTIC - ERROR
>> [1]  + Done                          ( cd $PWD; $t $exe
>> ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>> **  OPTIC crashed!
>> 0.036u 0.024s 0:05.26 0.9%0+0k 0+1016io 0pf+0w
>> error: command   /home/santu/wien2k/14/opticcpara -up -c -so upoptic.def
>>    failed
>>
>>
>> That is why I moved to normal GGA+U calculation and got again problem in
>>   optic and posted in forum.
>>
>> I would like to mention here when I did not add line "XMCD 1 L23" in
>> case.inop optic program ran without any problem.
>>
>>
>> So please suggest me to know the problem here and solve it.
>>
>> Thankin you,
>>
>> Santu Baidya
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On 15 May 2015 at 11:59, Peter Blaha <pblaha at theochem.tuwien.ac.at
>> <mailto:pblaha at theochem.tuwien.ac.at>> wrote:
>>
>>     You need spin-orbit coupling for   xmcd
>>
>>     Am 14.05.2015 um 20:58 schrieb Santu Baidya:
>>
>>         Dear Prof. Blaha and Wien2k users,
>>
>>              I am using wien2k version WIEN2k_14.1 (Release 8/9/2014)
>>         for calculating xmcd spectra of Co L23 edge for 8 atoms system.
>>         So I generated case.struct file. The did normal
>>         GGA+U spin polarized calculation using "runsp_lapw -orb -p".
>>
>>         Then I ran few commands as mentioned in userguide and wien2k 
>> forum:
>>         1) x kgen -p (for denser grid)
>>         2) x lapw1 -orb -up/dn -p (for eigenvalue calculations)
>>         3) x lapw2 -fermi -up/dn -p (for fermi energy calculation)
>>         4) editing case.inop file with additional line "XMCD 1 L23"
>>         5) x optic -orb -up -p
>>
>>         Before optic it ran very well. But after running optic it
>>         crashes with error:
>>
>>         running OPTIC in parallel mode
>>         [1] 13265
>>         [1]  + Done                          ( cd $PWD; $t $exe
>>         ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>>         [1] 13273
>>         [1]  + Done                          ( cd $PWD; $t $exe
>>         ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>>         [1] 13281
>>         [1]  + Done                          ( cd $PWD; $t $exe
>>         ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>>         [1] 13289
>>         [1]  + Done                          ( cd $PWD; $t $exe
>>         ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>>         [1] 13295
>>         [1]  + Done                          ( cd $PWD; $t $exe
>>         ${def}_${loop}.def; rm -f .lock_$lockfile[$p] ) >>  ...
>>         **  OPTIC crashed!
>>         0.016u 0.032s 0:05.26 0.7%0+0k 0+840io 0pf+0w
>>         error: command   /home/santu/wien2k/14/opticcpara -up -c
>>         upoptic.def   failed
>>
>>         After looking into other such error in wen2k forum I do not find
>>         its solution which led me to report it in the forum and asking
>>         for solution if known. My system has only 8
>>         atoms which is not big. So why should it crash !!!!
>>         Even I tried in GGA+U+SO calculation I got same problem which
>>         led me to do normal GGA+U calculation for trial. But it gives me
>>         same "optic crash".
>>
>>
>>         I would like to request you to help me to solve this problem in
>>         running optic program. Is wien2k optic program is unstable for
>>         XMCD calculation ?!
>>
>>
>>         Thanking you in advance,
>>
>>         Santu Baidya
>>         University of Duisburg
>>         Germany