Re: [Wien] how to use optB88-vdW with wien2k-18.1 (fwd)

[Gavin Abo]

Section "4.5.13 Nonlocal van der Waals functionals" on page 58 in the 
WIEN2k 18.2 usersguide [1] has the switch "-nlvdw".


Is the switch "-innlvdw" in your command below okay?

[1] http://susi.theochem.tuwien.ac.at/reg_user/textbooks/usersguide.pdf

On 8/17/2018 3:42 PM, Dr. K. C. Bhamu wrote:


Dear Tran,


I am not getting "case.r2v_nlvdw" file in case dir even I have a 
correct case.in0 and case.innlvdw



My log file and other information:


case.in0--
TOT  EX_OPTB88 EC_LDA VX_OPTB88 VC_LDA 
(XC_PBE,XC_PBESOL,XC_WC,XC_MBJ,XC_REVTPSSS)

NR2V  IFFT  (R2V)
   90   90   90    2.00  1    min IFFT-parameters, enhancement factor, 
iprint



initialized with

init_lapw -b -vxc 5 -rkmax 7  -numk 1000




>   (min_lapw) options: -j run_lapw -ec 0.0001  -cc 0.001 -p -fc 2 -i 
100 -innlvdw

Sat Aug 18 02:55:32 IST 2018> (x) dstart -super
>   (min_lapw) recover inm-file & call job run_lapw -ec 0.0001  -cc 
0.001 -p -fc 2 -i 100 -innlvdw

>   (run_lapw) options: -ec 0.1 -cc 0.0001 -p -fc 0.05 -i 100 -innlvdw
Sat Aug 18 02:55:35 IST 2018> (x) lapw0 -p
Sat Aug 18 02:56:02 IST 2018> (x) lapw1 -p
Sat Aug 18 02:56:31 IST 2018> (x) lapw2 -p
Sat Aug 18 02:56:46 IST 2018> (x) sumpara -d
Sat Aug 18 02:56:47 IST 2018> (x) lcore
Sat Aug 18 02:56:48 IST 2018> (x) mixer
Sat Aug 18 02:56:50 IST 2018> (x) lapw0 -p




Could you please have look and advice me if I am doing any mistake or 
we may not get case.r2v_nlvdw file?


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

Re: [Wien] how to use optB88-vdW with wien2k-18.1 (fwd)

[Dr. K. C. Bhamu]

Dear Tran,


I am not getting "case.r2v_nlvdw" file in case dir even I have a correct
case.in0 and case.innlvdw


My log file and other information:


case.in0--
TOT  EX_OPTB88 EC_LDA VX_OPTB88 VC_LDA
(XC_PBE,XC_PBESOL,XC_WC,XC_MBJ,XC_REVTPSSS)
NR2V  IFFT  (R2V)
   90   90   902.00  1min IFFT-parameters, enhancement factor,
iprint


initialized with

init_lapw -b -vxc 5 -rkmax 7  -numk 1000




>   (min_lapw) options: -j run_lapw -ec 0.0001  -cc 0.001 -p -fc 2 -i 100
-innlvdw
Sat Aug 18 02:55:32 IST 2018> (x) dstart -super
>   (min_lapw) recover inm-file & call job run_lapw -ec 0.0001  -cc 0.001
-p -fc 2 -i 100 -innlvdw
>   (run_lapw) options: -ec 0.1 -cc 0.0001 -p -fc 0.05 -i 100 -innlvdw
Sat Aug 18 02:55:35 IST 2018> (x) lapw0 -p
Sat Aug 18 02:56:02 IST 2018> (x) lapw1 -p
Sat Aug 18 02:56:31 IST 2018> (x) lapw2 -p
Sat Aug 18 02:56:46 IST 2018> (x) sumpara -d
Sat Aug 18 02:56:47 IST 2018> (x) lcore
Sat Aug 18 02:56:48 IST 2018> (x) mixer
Sat Aug 18 02:56:50 IST 2018> (x) lapw0 -p




Could you please have look and advice me if I am doing any mistake or we
may not get case.r2v_nlvdw file?

regards
Bhamu



On Fri, Aug 17, 2018 at 1:51 PM, Peter Blaha 
wrote:

> Yes, you can change it, but then you are using NOT what is called in
> literature optB88-vdW, but the "Bhamu-B88 functional".
>
> If you want to use what is called "optB88", you need to follow the
> instructions in the UG.
>
> Am 17.08.2018 um 10:04 schrieb Dr. K. C. Bhamu:
>
>> aah,
>>
>> I got it, yes we can!!
>>
>>
>> thanks
>>
>>
>>
>>
>> On Fri, Aug 17, 2018 at 1:29 PM, Dr. K. C. Bhamu > > wrote:
>>
>> Hii Tran,
>>
>> Sorry to interrupt you again,
>>
>> I see in UG that I need to modify case.in0 with EX_OPTB88 *EC_LDA
>> *VX_OPTB88 *VC_LDA*
>>
>> I am dealing all other cases with PBE so should I change LDA to PBE
>> or optB88_vdw is run only with LDA so that I should not change
>> anything as in above EX/EC/VX/VC?
>> Page number 116-118 of UG does not say much clear about PBE.
>>
>>
>>
>> regards
>> K.C. Bhamu
>>
>> On Fri, Aug 17, 2018 at 12:49 AM, > > wrote:
>>
>> Hi,
>>
>> The file case.r2v_nlvdw will be generated and used during the
>> calculation. You don't need to care about it.
>>
>> The steps for DOS, band structure and optics are exactly the same
>> as with usual LDA or GGA.
>>
>> The extra computational time due to NLVDW does not depend on
>> RKMAX or k-mesh. It depends on the size of the unit cell and on
>> the value of plane-wave expansion cutoff GMAX in case.innlvdw.
>>
>> If you don't need to optimize position of atoms in the
>> unit cell (i.e., no "-min"), then replace "T" by "F" in the
>> last line of case.innlvdw. This will reduce significantly
>> the NLVDW computational time.
>>
>> F. Tran
>>
>> Dear Wien2k users
>>
>> I have a few questions for optB88-vdW with Wien2k_18.1.
>>
>>
>> I need to use optB88-vdW for a perovskite structure.
>>
>> What I found from the mailing list and UG;
>>
>> 1. Need two files case.in0 and case.innlvdw to use this
>> function and can optimize the structure with this NL
>> functional.
>> 2. One should add "–nlvdw" run(sp)_lapw script.
>>
>>
>> Now I have below queries:
>>
>> 1. It is mentioned to use "case.r2v_nlvdw" file [1], but in
>> UG nothing is said about this file. If we need to use
>> case.r2v_nlvdw then how to recall it?
>>
>> 2. Do we need to treat the step in the same way as we do for
>> doss, optical and band structure or we need to modify them? If
>> we need to modify these steps then what are the necessary
>> changes (x -h  show nothing about -nlvdw for calculating
>> these
>> properties)?
>>
>> 3. How expensive is in comparison to PBE? If I use 12x12x12
>> mesh for PBE then how much I can reduce the mesh size
>> (running on
>> 16 processor CPU)?
>>
>>
>> [1]
>> https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at
>> /msg16549.html
>> > t/msg16549.html>
>>
>>
>>
>> Thanks and regards
>>
>> K.C. Bhamu
>>
>>
>> ___
>> Wien mailing list
>> Wien@zeus.theochem.tuwien.ac.at
>> 
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>> 
>> SEARCH t

Re: [Wien] Error in Fermi surface for GGA+U+SOC calculations

[Gavin Abo]

I tried the files you sent me.

When I click on the "ErrorInfo" button, it gives me:

At line 125 of file wn_readbakgegn.f (unit = 8, file = 'case.outputkgen')
Fortran runtime error: Bad integer for item 1 in list input

Based on the XCrySDen mailing list post for that error at:

http://www.democritos.it/pipermail/xcrysden/2015-December/001788.html

It looks like the "weights of k-points" and sum of the "NUMBER OF 
K-POINTS" should match.


However, when I look at your case.outputkgen, it look like their are 126 
"weights of k-points".


Your case.output1up seems to have 152 when the multiple "NUMBER OF 
K-POINTS" lines are summed together (instead of the matching 126).


Maybe more case.outputso_* files were produced at some point in your 
calculation.  So you maybe doing a cat of too many case.outputso_* files.


For example, if you had case.outputso_1 ... case.outputso_10 giving 
152.  It may be that only case.outputso_1 ... case.outputso_8 should be 
combined using cat to get 126 in case.output1up.  The case.outputso_9 
and case.outputso_10 giving a sum of 26 for "NUMBER OF K-POINTS" may be 
from some other calculation at some point.


On 8/13/2018 11:14 AM, Anup Shakya wrote:

Dear Prof. Blaha,

Sorry to disturb you again. I have followed all the steps as suggested 
by you.


x lapw1 -up -p
x lapw1 -dn -p
x lapwso -orb -up -p

I have concatenated all the case.output files into a single file 
case.output1up
then performed x lapw2 -so -fermi -up -p, opened xcrysden and clicked 
Render Fermi surface. I still get an error


ERROR: while executing exec/usr/lib/xcrysden/wn_readbakgen bakgen.def

I created a new directory and performed the calculations from scratch 
again but still the error persists. I have done non magnetic GGA+SOC+U 
calculations. Does the command change for this type of calculations.



I would be grateful to you if you could suggest me where I went wrong. 
I did search the previous posts but its not helpful for this problem.


Anup Pradhan Sakhya


___
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] problem with YS-PBE0

[Dr. K. C. Bhamu]

Dear Wienk Users,

I attempted to do a band structure calculation of a perovskite structure
with YS-PBE0 (standard alpa parameter) with Wien2k-18.1.
Up to scf and doss, I do not see any problem.

But I am not getting optical properties and below is what I am getting in
band.agr file:


My log file is:

Thu Aug 16 17:15:58 IST 2018> (x_lapw) lapw1 -p
Thu Aug 16 17:16:24 IST 2018> (x_lapw) lapw2 -qtl -p -hf
Thu Aug 16 17:16:29 IST 2018> (x_lapw) tetra -p -hf
Thu Aug 16 17:16:40 IST 2018> (x_lapw) lapw1 -p
Thu Aug 16 17:17:08 IST 2018> (x_lapw) lapw2 -fermi -p -hf
Thu Aug 16 17:17:11 IST 2018> (x_lapw) optic -p -hf
Thu Aug 16 17:17:20 IST 2018> (x_lapw) joint -p -hf
Thu Aug 16 17:17:31 IST 2018> (x_lapw) kram -p
Thu Aug 16 17:17:46 IST 2018> (x_lapw) lapw1 -band -p
Thu Aug 16 17:29:48 IST 2018> (x_lapw) lapw2 -band -qtl -p -hf
Thu Aug 16 17:29:52 IST 2018> (x_lapw) spaghetti -p -hf

I did not increase the k-points for bands and optical properties and
continued with 3x3x3 mesh (total 4 points).

@ xaxis  tick major   0, 0.0
 @ xaxis  ticklabel0 ," 1  "
@ xaxis  tick major   1, 0.15766
 @ xaxis  ticklabel1 ," 2  "
@ xaxis  tick major   2, 0.31533
 @ xaxis  ticklabel2 ," 3  "
@ xaxis  tick major   3, 0.47299
 @ xaxis  ticklabel3 ," 4  "
@ with g0
@ world 0,-21.0, 0.47299,12.0
 @ autoticks
 @ yaxis  label "Energy(eV)"
 @ with line
 @ line on
 @ line loctype world
@ line  0.0, 0.0, 0.47299, 0.0
 @ line linestyle 3
 @ line def
 @ with string
 @ string on
 @ string loctype world
@ string  0.49299, -0.1
 @ string char size 1.50
 @ string def "E\sF"
 @ title "CSI_ysfc"
 #k   ene character
 @ autoscale onread none
 @ target g0.s1
 @ type xysize

 # bandindex:   1
   0.0 -69.93183   0.07000
   0.15766 -69.93147   0.07000
   0.31533 -69.93165   0.07000
   0.47299 -69.93150   0.07000
&
 @ autoscale onread none
 @ target g0.s2
 @ type xysize

 # bandindex:   2
   0.0 -69.93183   0.07000
   0.15766 -69.93147   0.07000
   0.31533 -69.93151   0.07000
   0.47299 -69.93149   0.07000

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

Re: [Wien] how to use optB88-vdW with wien2k-18.1 (fwd)

[Dr. K. C. Bhamu]

Thanks Prof. Peter,

I got the point now.

Regards
K.C. Bhamu




On Fri, Aug 17, 2018 at 1:51 PM, Peter Blaha 
wrote:

> Yes, you can change it, but then you are using NOT what is called in
> literature optB88-vdW, but the "Bhamu-B88 functional".
>
> If you want to use what is called "optB88", you need to follow the
> instructions in the UG.
>
> Am 17.08.2018 um 10:04 schrieb Dr. K. C. Bhamu:
>
>> aah,
>>
>> I got it, yes we can!!
>>
>>
>> thanks
>>
>>
>>
>>
>> On Fri, Aug 17, 2018 at 1:29 PM, Dr. K. C. Bhamu > > wrote:
>>
>> Hii Tran,
>>
>> Sorry to interrupt you again,
>>
>> I see in UG that I need to modify case.in0 with EX_OPTB88 *EC_LDA
>> *VX_OPTB88 *VC_LDA*
>>
>> I am dealing all other cases with PBE so should I change LDA to PBE
>> or optB88_vdw is run only with LDA so that I should not change
>> anything as in above EX/EC/VX/VC?
>> Page number 116-118 of UG does not say much clear about PBE.
>>
>>
>>
>> regards
>> K.C. Bhamu
>>
>> On Fri, Aug 17, 2018 at 12:49 AM, > > wrote:
>>
>> Hi,
>>
>> The file case.r2v_nlvdw will be generated and used during the
>> calculation. You don't need to care about it.
>>
>> The steps for DOS, band structure and optics are exactly the same
>> as with usual LDA or GGA.
>>
>> The extra computational time due to NLVDW does not depend on
>> RKMAX or k-mesh. It depends on the size of the unit cell and on
>> the value of plane-wave expansion cutoff GMAX in case.innlvdw.
>>
>> If you don't need to optimize position of atoms in the
>> unit cell (i.e., no "-min"), then replace "T" by "F" in the
>> last line of case.innlvdw. This will reduce significantly
>> the NLVDW computational time.
>>
>> F. Tran
>>
>> Dear Wien2k users
>>
>> I have a few questions for optB88-vdW with Wien2k_18.1.
>>
>>
>> I need to use optB88-vdW for a perovskite structure.
>>
>> What I found from the mailing list and UG;
>>
>> 1. Need two files case.in0 and case.innlvdw to use this
>> function and can optimize the structure with this NL
>> functional.
>> 2. One should add "–nlvdw" run(sp)_lapw script.
>>
>>
>> Now I have below queries:
>>
>> 1. It is mentioned to use "case.r2v_nlvdw" file [1], but in
>> UG nothing is said about this file. If we need to use
>> case.r2v_nlvdw then how to recall it?
>>
>> 2. Do we need to treat the step in the same way as we do for
>> doss, optical and band structure or we need to modify them? If
>> we need to modify these steps then what are the necessary
>> changes (x -h  show nothing about -nlvdw for calculating
>> these
>> properties)?
>>
>> 3. How expensive is in comparison to PBE? If I use 12x12x12
>> mesh for PBE then how much I can reduce the mesh size
>> (running on
>> 16 processor CPU)?
>>
>>
>> [1]
>> https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at
>> /msg16549.html
>> > t/msg16549.html>
>>
>>
>>
>> Thanks and regards
>>
>> K.C. 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
>> > /index.html>
>>
>>
>>
>>
>>
>> ___
>> 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/wi
>> e...@zeus.theochem.tuwien.ac.at/index.html
>>
>>
> --
> --
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
> Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
> WWW: http://www.imc.tuwien.ac.at/tc_blaha
> -
> ___
> 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/wi
> e...@zeus.theochem.tuwien.ac.at/index.html
>
___
Wien 

Re: [Wien] how to use optB88-vdW with wien2k-18.1 (fwd)

[Peter Blaha]

Yes, you can change it, but then you are using NOT what is called in 
literature optB88-vdW, but the "Bhamu-B88 functional".


If you want to use what is called "optB88", you need to follow the 
instructions in the UG.


Am 17.08.2018 um 10:04 schrieb Dr. K. C. Bhamu:

aah,

I got it, yes we can!!


thanks




On Fri, Aug 17, 2018 at 1:29 PM, Dr. K. C. Bhamu > wrote:


Hii Tran,

Sorry to interrupt you again,

I see in UG that I need to modify case.in0 with EX_OPTB88 *EC_LDA
*VX_OPTB88 *VC_LDA*

I am dealing all other cases with PBE so should I change LDA to PBE
or optB88_vdw is run only with LDA so that I should not change
anything as in above EX/EC/VX/VC?
Page number 116-118 of UG does not say much clear about PBE.



regards
K.C. Bhamu

On Fri, Aug 17, 2018 at 12:49 AM, mailto:t...@theochem.tuwien.ac.at>> wrote:

Hi,

The file case.r2v_nlvdw will be generated and used during the
calculation. You don't need to care about it.

The steps for DOS, band structure and optics are exactly the same
as with usual LDA or GGA.

The extra computational time due to NLVDW does not depend on
RKMAX or k-mesh. It depends on the size of the unit cell and on
the value of plane-wave expansion cutoff GMAX in case.innlvdw.

If you don't need to optimize position of atoms in the
unit cell (i.e., no "-min"), then replace "T" by "F" in the
last line of case.innlvdw. This will reduce significantly
the NLVDW computational time.

F. Tran

Dear Wien2k users

I have a few questions for optB88-vdW with Wien2k_18.1.


I need to use optB88-vdW for a perovskite structure.

What I found from the mailing list and UG;

1. Need two files case.in0 and case.innlvdw to use this
function and can optimize the structure with this NL functional.
2. One should add "–nlvdw" run(sp)_lapw script.


Now I have below queries:

1. It is mentioned to use "case.r2v_nlvdw" file [1], but in
UG nothing is said about this file. If we need to use
case.r2v_nlvdw then how to recall it?

2. Do we need to treat the step in the same way as we do for
doss, optical and band structure or we need to modify them? If
we need to modify these steps then what are the necessary
changes (x -h  show nothing about -nlvdw for calculating
these
properties)?

3. How expensive is in comparison to PBE? If I use 12x12x12
mesh for PBE then how much I can reduce the mesh size
(running on
16 processor CPU)?


[1]

https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg16549.html





Thanks and regards

K.C. 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






___
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



--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW: 
http://www.imc.tuwien.ac.at/tc_blaha- 


___
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

Re: [Wien] how to use optB88-vdW with wien2k-18.1 (fwd)

[Dr. K. C. Bhamu]

aah,

I got it, yes we can!!


thanks




On Fri, Aug 17, 2018 at 1:29 PM, Dr. K. C. Bhamu 
wrote:

> Hii Tran,
>
> Sorry to interrupt you again,
>
> I see in UG that I need to modify case.in0 with EX_OPTB88 *EC_LDA *VX_OPTB88
> *VC_LDA*
>
> I am dealing all other cases with PBE so should I change LDA to PBE or
> optB88_vdw is run only with LDA so that I should not change anything as in
> above EX/EC/VX/VC?
> Page number 116-118 of UG does not say much clear about PBE.
>
>
>
> regards
> K.C. Bhamu
>
> On Fri, Aug 17, 2018 at 12:49 AM,  wrote:
>
>> Hi,
>>
>> The file case.r2v_nlvdw will be generated and used during the
>> calculation. You don't need to care about it.
>>
>> The steps for DOS, band structure and optics are exactly the same
>> as with usual LDA or GGA.
>>
>> The extra computational time due to NLVDW does not depend on
>> RKMAX or k-mesh. It depends on the size of the unit cell and on
>> the value of plane-wave expansion cutoff GMAX in case.innlvdw.
>>
>> If you don't need to optimize position of atoms in the
>> unit cell (i.e., no "-min"), then replace "T" by "F" in the
>> last line of case.innlvdw. This will reduce significantly
>> the NLVDW computational time.
>>
>> F. Tran
>>
>> Dear Wien2k users
>>>
>>> I have a few questions for optB88-vdW with Wien2k_18.1.
>>>
>>>
>>> I need to use optB88-vdW for a perovskite structure.
>>>
>>> What I found from the mailing list and UG;
>>>
>>> 1. Need two files case.in0 and case.innlvdw to use this function and can
>>> optimize the structure with this NL functional.
>>> 2. One should add "–nlvdw" run(sp)_lapw script.
>>>
>>>
>>> Now I have below queries:
>>>
>>> 1. It is mentioned to use "case.r2v_nlvdw" file [1], but in UG nothing
>>> is said about this file. If we need to use
>>> case.r2v_nlvdw then how to recall it?
>>>
>>> 2. Do we need to treat the step in the same way as we do for doss,
>>> optical and band structure or we need to modify them? If
>>> we need to modify these steps then what are the necessary changes (x -h
>>>  show nothing about -nlvdw for calculating these
>>> properties)?
>>>
>>> 3. How expensive is in comparison to PBE? If I use 12x12x12 mesh for PBE
>>> then how much I can reduce the mesh size (running on
>>> 16 processor CPU)?
>>>
>>>
>>> [1] https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at
>>> /msg16549.html
>>>
>>>
>>>
>>> Thanks and regards
>>>
>>> K.C. 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/wi
>> e...@zeus.theochem.tuwien.ac.at/index.html
>>
>>
>
___
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

Re: [Wien] how to use optB88-vdW with wien2k-18.1 (fwd)

[Dr. K. C. Bhamu]

Hii Tran,

Sorry to interrupt you again,

I see in UG that I need to modify case.in0 with EX_OPTB88 *EC_LDA *VX_OPTB88
*VC_LDA*

I am dealing all other cases with PBE so should I change LDA to PBE or
optB88_vdw is run only with LDA so that I should not change anything as in
above EX/EC/VX/VC?
Page number 116-118 of UG does not say much clear about PBE.



regards
K.C. Bhamu

On Fri, Aug 17, 2018 at 12:49 AM,  wrote:

> Hi,
>
> The file case.r2v_nlvdw will be generated and used during the
> calculation. You don't need to care about it.
>
> The steps for DOS, band structure and optics are exactly the same
> as with usual LDA or GGA.
>
> The extra computational time due to NLVDW does not depend on
> RKMAX or k-mesh. It depends on the size of the unit cell and on
> the value of plane-wave expansion cutoff GMAX in case.innlvdw.
>
> If you don't need to optimize position of atoms in the
> unit cell (i.e., no "-min"), then replace "T" by "F" in the
> last line of case.innlvdw. This will reduce significantly
> the NLVDW computational time.
>
> F. Tran
>
> Dear Wien2k users
>>
>> I have a few questions for optB88-vdW with Wien2k_18.1.
>>
>>
>> I need to use optB88-vdW for a perovskite structure.
>>
>> What I found from the mailing list and UG;
>>
>> 1. Need two files case.in0 and case.innlvdw to use this function and can
>> optimize the structure with this NL functional.
>> 2. One should add "–nlvdw" run(sp)_lapw script.
>>
>>
>> Now I have below queries:
>>
>> 1. It is mentioned to use "case.r2v_nlvdw" file [1], but in UG nothing is
>> said about this file. If we need to use
>> case.r2v_nlvdw then how to recall it?
>>
>> 2. Do we need to treat the step in the same way as we do for doss,
>> optical and band structure or we need to modify them? If
>> we need to modify these steps then what are the necessary changes (x -h
>>  show nothing about -nlvdw for calculating these
>> properties)?
>>
>> 3. How expensive is in comparison to PBE? If I use 12x12x12 mesh for PBE
>> then how much I can reduce the mesh size (running on
>> 16 processor CPU)?
>>
>>
>> [1] https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at
>> /msg16549.html
>>
>>
>>
>> Thanks and regards
>>
>> K.C. 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
>
>
___
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

Re: [Wien] Error in Fermi surface for GGA+U+SOC calculations

[Gavin Abo]

Does your structure for this SO+U case allow a shifted k-mesh [1]? If 
so, make sure during "x kgen" that you selected the non-shifted k-mesh 
as the XCrySDen website [2] says:


"File-->Open WIEN2k ...-->Fermi Surface Pops-up a task window for Fermi 
surface creation. After several steps the Fermi surface is hopefully 
drawn as 3D isosurface. This feature is EXPERIMENTAL, please be careful 
!!! So far it was tested on a few spin non-polarized and spin-polarized 
systems. (Currently the *shift of the k-mesh is not allowed*.)"


As I recall, serial case.outputso file used with the XCrySDen fermi 
surface program ran fine [3].  So likely it will run fine too when the 
parallel outputso files are combined:


cat case.outputso_1 case.outputso_2 ... > case.output1up

Previously, you mentioned that SO (without U) ran fine.  I would 
anticipate differences in values in the files, but likely formatting of 
the files would be the same.  Have you compared the formatting of 
case.output1up file between the SO (without U) and SO+U calculations.  
Any differences?


XCrySDen 1.5.60 might have to patched if you are using too many k-points 
or bands [5, 6].


Any error messages in the file case.outputbaup?

If you need further help, can you send a link to the SO+U files for the 
case below to the mailing list (or to my personal email)?  I believe at 
least the files case.struct, case.klist, case.clmdn, case.outputkgen, 
and case.output1up are needed to open and execute "Render Fermi Surface" 
in XCrySDen.


[1] http://susi.theochem.tuwien.ac.at/reg_user/faq/kgen.html
[2] http://www.xcrysden.org/doc/wien.html#fermi
[3] 
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg02058.html
[4] 
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg17831.html
[5] 
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15219.html

[6] http://www.democritos.it/pipermail/xcrysden/2015-October/001753.html

On 8/13/2018 11:14 AM, Anup Shakya wrote:

Dear Prof. Blaha,

Sorry to disturb you again. I have followed all the steps as suggested 
by you.


x lapw1 -up -p
x lapw1 -dn -p
x lapwso -orb -up -p

I have concatenated all the case.output files into a single file 
case.output1up
then performed x lapw2 -so -fermi -up -p, opened xcrysden and clicked 
Render Fermi surface. I still get an error


ERROR: while executing exec/usr/lib/xcrysden/wn_readbakgen bakgen.def

I created a new directory and performed the calculations from scratch 
again but still the error persists. I have done non magnetic GGA+SOC+U 
calculations. Does the command change for this type of calculations.



I would be grateful to you if you could suggest me where I went wrong. 
I did search the previous posts but its not helpful for this problem.


Anup Pradhan Sakhya
___
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