Re: [Wien] Convergence problem within the method GGA+U

[Abderrahmane Reggad]

Thanks again Dr Fecher for your answers

- it can never stop precisely at the number you give

That's what happened in my case but according to the statement of Prof Blaha

we can't refer to the web screen printing and we have to check the
case.daytime during every cycle

Best regards



-- 
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie
___
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] Convergence problem within the method GGA+U

[Abderrahmane Reggad]

Thank you Prof Blaha for your answer


1- "From the file on the web, you cannot determine the final convergence"

 I haven't been knowing that before and thanks before for this
information and I realize

that's the source of confusion.

2- "The numbers you quote are for the last but one iteration, the
convergence test AFTER the last iteration is not printed."

I have presented the last iteration of each cycle representing defined structure

"- "This is printed only in the case.dayfile "

Unfortunately it's the last cycle that's printed in this file

I will repeat the calculation and check the convergence during every cycle


PS: I would be more concerned where the unmatched"   comes ??

I have already presented this before and I didn't received a full  answer

This problem appears always when I use the command line " runsp_c_lapw
" but it doesn't influence the calculation

Best regards

-- 
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie
___
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] w2w compilation problem during siteconfig

[Fecher, Gerhard]

Hallo Elias,
thanks for the answers.

Meanwhile I checked, the error does not appear with GNU make 4.0
seems we should update our Cluster a little bit, I have to check which of our 
Computers are affected.

Personally I like more to tell make which files I like to have compiled instead 
of using an automatic 
search and the makefile decides which ones I like to have compiled, 
but maybe I am just a little bit too old fashioned.

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-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Elias 
Assmann [elias.assm...@gmail.com]
Gesendet: Mittwoch, 21. Dezember 2016 19:21
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] w2w compilation problem during siteconfig

My memory may be feeble, but the Internet remembers All.

It turns out that I had this exact problem three years ago.  I even
asked a question about it on StackOverflow [1], where a helpful user
pointed out that this is a known bug in make 3.82 [2] and suggested the
solution

.SECONDEXPANSION:

%.o: $$(wildcard %.[fF])
$(FC) $(FFLAGS) -c $< -o $@

(splitting the rule should still also work, of course).

Since this problem was fixed in later versions of make and is easy to
circumvent if you are affected, I will leave my Makefiles as they are.

It remains a puzzle though that I cannot now reproduce the problem with
make 3.82 (neither with a make 3.81 that I found elsewhere).  Maybe
RedHat backported the fix?  ‘make -v’ says

GNU Make 3.82
Built for x86_64-redhat-linux-gnu


Elias


[1] http://stackoverflow.com/questions/21193024/wildcard-in-a-pattern-rule
[2] https://savannah.gnu.org/bugs/?func=detailitem_id=31621

--
Elias Assmann
Institute of Theoretical and Computational Physics
TU Graz   

___
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] Convergence problem within the method GGA+U

[Fecher, Gerhard]

I think you got that completely wrong
the criterion means that the scf cycle stops when the values are better than 
those given in ec and /or cc
it can never stop precisely at the number you give

However, you never told which values you gave and which ones were reached, so 
what ?

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-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Abderrahmane 
Reggad [jazai...@gmail.com]
Gesendet: Donnerstag, 22. Dezember 2016 21:08
An: wien@zeus.theochem.tuwien.ac.at
Betreff: Re: [Wien] Convergence problem within the method GGA+U

Hello again

When you choose a convergence criterion for energy equal to 0.1 , do you 
think it's normal that the scf cycle stops arbitrarily ; onetime stops when dE= 
0.59 and another time stops when dE= 0.23 and so on.

I didn't see such a situation in the previous versions

Best regards


--
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie


___
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] Convergence problem within the method GGA+U

[Fecher, Gerhard]

I was just in a hurry, to be more precise,
why did you delete those lines ?
How did they influence the convergence ?
what other lines did you change in the script ?

To just delete those lines does not make sense

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-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Fecher, 
Gerhard [fec...@uni-mainz.de]
Gesendet: Donnerstag, 22. Dezember 2016 20:36
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] Convergence problem within the method GGA+U

why ? how ? for what ?

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-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Abderrahmane 
Reggad [jazai...@gmail.com]
Gesendet: Donnerstag, 22. Dezember 2016 19:01
An: wien@zeus.theochem.tuwien.ac.at
Betreff: Re: [Wien] Convergence problem within the method GGA+U

Hello again

I have deleted the following option of charge extrapolation because with it the 
convergence was worst

clmextrapol_lapw
 if (-e NiS-afmI-optV_U0-ec1.clmup &&  ! -z NiS-afmI-optV_U0-ec1.clmup) 
then
 clmextrapol_lapw -up
 clmextrapol_lapw -dn
 endif

Best regards

--
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie


___
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] Convergence problem within the method GGA+U

[Abderrahmane Reggad]

Hello again

When you choose a convergence criterion for energy equal to 0.1 , do
you think it's normal that the scf cycle stops arbitrarily ; onetime stops
when dE= 0.59 and another time stops when dE= 0.23 and so on.

I didn't see such a situation in the previous versions

Best regards



-- 
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie
___
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] Convergence problem within the method GGA+U

[Fecher, Gerhard]

why ? how ? for what ?

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-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Abderrahmane 
Reggad [jazai...@gmail.com]
Gesendet: Donnerstag, 22. Dezember 2016 19:01
An: wien@zeus.theochem.tuwien.ac.at
Betreff: Re: [Wien] Convergence problem within the method GGA+U

Hello again

I have deleted the following option of charge extrapolation because with it the 
convergence was worst

clmextrapol_lapw
 if (-e NiS-afmI-optV_U0-ec1.clmup &&  ! -z NiS-afmI-optV_U0-ec1.clmup) 
then
 clmextrapol_lapw -up
 clmextrapol_lapw -dn
 endif

Best regards

--
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie


___
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] Convergence problem within the method GGA+U

[Fecher, Gerhard]

what is your problem ?
your file tells that all cases are converged within about 10 cycles and in non 
of the cases the number of scf cycles was exhausted.

where are those seven numbers coming from, I do not remember that Wien2k has 
such an output
and a last time, without knowing your input no one knows what you did

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-boun...@zeus.theochem.tuwien.ac.at] im Auftrag von Abderrahmane 
Reggad [jazai...@gmail.com]
Gesendet: Donnerstag, 22. Dezember 2016 16:56
An: wien@zeus.theochem.tuwien.ac.at
Betreff: [Wien] Convergence problem within the method GGA+U

Dear wien2k users

To determine to magnetic ground state among three states , I have made three 
calculations for each state .

The problem encontred is to get the convergence  for all the case even with the 
increade of k points number and the energy convergence criteria.

I have joined the SCF cycle for ec=1 in the following link

https://docs.google.com/document/d/1LGQvsWUOcW-oYnBt1QzEZcKPNfdT1mecVPn5XNhrMjc/edit?usp=sharing

Here is a summary of result:

-6> 0.59
-3> 0.23
0 > 0.18
3 > 0.15
6 > 0.20
9 > 0.15


Best regards

--
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie


___
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] Convergence problem within the method GGA+U

[Abderrahmane Reggad]

Hello again

I have deleted the following option of charge extrapolation because with it
the convergence was worst

clmextrapol_lapw
 if (-e NiS-afmI-optV_U0-ec1.clmup &&  ! -z
NiS-afmI-optV_U0-ec1.clmup) then
 clmextrapol_lapw -up
 clmextrapol_lapw -dn
 endif

Best regards

-- 
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie
___
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] Convergence problem within the method GGA+U

[Abderrahmane Reggad]

Dear wien2k users

To determine to magnetic ground state among three states , I have made
three calculations for each state .

The problem encontred is to get the convergence  for all the case even with
the increade of k points number and the energy convergence criteria.

I have joined the SCF cycle for ec=1 in the following link

https://docs.google.com/document/d/1LGQvsWUOcW-oYnBt1QzEZcKPNfdT1mecVPn5XNhrMjc/edit?usp=sharing

Here is a summary of result:

-6> 0.59
-3> 0.23
0 > 0.18
3 > 0.15
6 > 0.20
9 > 0.15


Best regards

-- 
Mr: A.Reggad
Laboratoire de Génie Physique
Université Ibn Khaldoun - Tiaret
Adresse: BP 144 AL ATTAF AIN DEFLA
Tel: +213(0)561861963
Algerie
___
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] Energy for optical.properties

[Karel Vyborny]

Dear ben amara imen,

my amateur opinion is that you should be very careful even when using 
calculated optical response for energies beyond 10 eV. Bands higher than 
several eV above the Fermi level lie in continuum.


Best regards,

KV


--- x ---
dr. Karel Vyborny
Fyzikalni ustav AV CR, v.v.i.
Cukrovarnicka 10
Praha 6, CZ-16253
tel: +420220318459


On Thu, 22 Dec 2016, ben amara imen wrote:



Hi

Im calculating the dielectric function and Eloss in [0-15 eV]  using wien2k
My question is : can I extend my calculation to 30 eV ? And how ?

Thanks in advance


___
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] Energy for optical.properties

[ben amara imen]

Hi

Im calculating the dielectric function and Eloss in [0-15 eV]  using wien2k
My question is : can I extend my calculation to 30 eV ? And how ?

Thanks in advance
___
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