SCM

help with densf program

shantanu — Fri Aug 15, 2008 5:19 am

Author: shantanu

Posted: Fri Aug 15, 2008 5:19 am

hi marcel,
Thanks for your reply. It really helped. I generated two tape41 files
with the option ALL HOMO-5 LUMO +5.
For some reason HOMO orbital was missing from the list for both of them.
My understanding is that with the above command all orbitals in the range are calculated. Is that right?
Thanks
Shantanu

marcelswart wrote:

You can call it directly from the command line as

Code:

$ADFBIN/densf < input

but you should already have a TAPE21 file present.

It's the same for with using SGE, but then it depends on
your settings (e.g. if you start from your home directory,
or from the directory where you submitted the job, or from
the $TMPDIR), so in that case you'd have to make sure that
the TAPE21 file is present in the working directory,
and probably you will need to put the complete path of
the inputfile

Code:

$ADFBIN/densf < /here/the/complete/path/input

UNRESTRICTED FRAGMENT error

jasius — Thu Aug 14, 2008 9:05 pm

Author: jasius

Posted: Thu Aug 14, 2008 9:05 pm

anybody?..

parallization on dual core processor, Help from ADF users

may01dz — Fri Aug 01, 2008 9:16 am

Author: may01dz

Posted: Fri Aug 01, 2008 9:16 am

dear ADF'ers,
I have the ADF2007.01 version (the parallel one: in an HP-MPI runtime), and, I would like to know, step by step, how to configure my ADF code to work in parallel manner on one computer but have a dual core processor.
Thank you for your time and any assistance you can provide

ADF-GUI

marcelswart — Tue Jul 29, 2008 6:56 am

Author: marcelswart

Posted: Tue Jul 29, 2008 6:56 am

Choose the "View" menu, "Background" item, and set it to white.

ADF crashes machine (but it never used to)

Anna Anastasi — Mon Jul 28, 2008 11:31 am

Author: Anna Anastasi

Posted: Mon Jul 28, 2008 11:31 am

Dear Alexei,

Thanks for you reply. See below for processor details. Since upgrading the operating system (to from Linux Suse Enterprise Server 10 SP1 (x86_64) to Linux Suse Enterprise 11 the jobs still crash but run for a bit, and they don't make the computer unresponsive. However, they don't run to completion.

The specs of the computer (aside from those in the original post) are:
> - Two dual core Intel Xeon
> - 4GB main memory
> - (glibc version: 2.4)

Memory wise, we did a memory test which didn't highlight any memory problems, and there is 4GB of swap memory:
total used free shared buffers cached
Mem: 4049056 516668 3532388 0 26064 393360
-/+ buffers/cache: 97244 3951812
Swap: 4200956 0 4200956

Job wise, I have tried a couple of systems but the main one I have been looking at has 24 Atoms including 2 Rh, 10O, 4C, and uses scalar relativistic effects, and COSMO. Initially (under Suse ... 10) I hadn't specified any memory limits but now vectorlength is set to 64 and IObuffer is set to 128. I haven't tried increasing these yet. There is about 70GB free in the /scratch directory where the scratch files are.

I wondered whether there was some problem in directing the jobs to the correct libraries given that the code wasn't specifically compiled for our system? We are using the pc_linux_64 binary at the moment.

Thanks in advance for your help.

Anna

TD-DFT

shantanu — Mon Jul 28, 2008 5:38 am

Author: shantanu

Posted: Mon Jul 28, 2008 5:38 am

I would appreciate if someone can explain to me, how TD-DFT becomes applicable to describing excited state electronic states. in other words how is an excited electronic state related to the time-dependedence of perturbation.
I would also appreciate if you could cite some research papers .
Thanks
Shantanu

parallel calculation on dual core processors

may01dz — Sat Jul 26, 2008 8:16 pm

Author: may01dz

Posted: Sat Jul 26, 2008 8:16 pm

dear ADF'ers
it is any possibility to run parallel calculation on a computure have a dual core processor.
thanks

openshell TDDFT energies

vladimir — Sat Jul 05, 2008 5:09 pm

Author: vladimir

Posted: Sat Jul 05, 2008 5:09 pm

When doing unrestricted calculation then you *should* get slightly different energies for alpha and beta electrons, if you are having an open shell system
(a system where you expect alpha and beta spin densities to differ from each other; and I guess this would be true for Mn(III) porphyrin).
This phenomenon is sometimes called spin polarization.

However if you choose a closed shell system (for example a H2O molecule) then using unrestricted formalism will give you exactly the same alpha and beta electron energies within the same orbital.
But for a closed shell system, using unrestricted formalism would be a waste of computational time.
(UNRESTRICTED keyword builds separate solutions for alpha and beta electrons, and therefore including this keyword will increase the computational time).

Excitation of open shell systems (TDDFT)

yuki — Fri May 23, 2008 4:18 am

Author: yuki

Posted: Fri May 23, 2008 4:18 am

I want to know if ADF can calculate the hyperpolarization of the open shell system?

error using densf program

Alexei Yakovlev — Thu May 08, 2008 1:27 pm

Author: Alexei Yakovlev

Posted: Thu May 08, 2008 1:27 pm

Hi Shantanu,

Please send the complete output file to me. You should know my email address.

Alexei

Points file is it good for anything ?

nuno — Tue May 06, 2008 2:57 pm

Author: nuno

Posted: Tue May 06, 2008 2:57 pm

When my Band job is finished I get a binary file called Points. What's it for?

Dipole Moment of The Excited State

marcelswart — Wed Apr 30, 2008 9:25 am

Author: marcelswart

Posted: Wed Apr 30, 2008 9:25 am

See the discussion about "a question about excited state dipole moment".
Yes, you can with ADF if you use finite field differences.

See for instance:

Quote:

F.C.Grozema, R.Telesca, H.T.Jonkman, L. Siebeles, J.G.Snijders
Excited State Polarizabilities of Conjugated Molecules Calculated Using Time Dependent Density Functional Theory
J. Chem. Phys. 2001, 115, 10014-10021

Although they used the methodology to compute the polarizability of the excited state, this is in fact one step more difficult because this involves the *derivative* of the dipole moment of the excited state.

a question about excited state dipole moment

marcelswart — Wed Apr 30, 2008 9:23 am

Author: marcelswart

Posted: Wed Apr 30, 2008 9:23 am

See for instance:

Quote:

error in excitation energy calculation

marcelswart — Tue Apr 29, 2008 7:19 am

Author: marcelswart

Posted: Tue Apr 29, 2008 7:19 am

Excitations are within the standard excitations framework only possible with closed-shell molecules, see
http://www.scm.com/Doc/Doc2007.01/ADF/ADFUsersGuide/page119.html#keyscheme%20EXCITATIONS.

For unrestricted molecules like yours, you should have a look at the Spin-Flip TD-DFT implementation by Ziegler and co-workers:
http://www.scm.com/Doc/Doc2007.01/ADF/ADFUsersGuide/page121.html#keyscheme%20SFTDDFT.

densf program

shantanu — Mon Apr 28, 2008 7:43 pm

Author: shantanu

Posted: Mon Apr 28, 2008 7:43 pm

does any one know the syntax for running densf command on the command promt ?