My guess regarding the compilation errors: there's no need to use sudo for this install, and your compiler in the makefile is pointing to f77, is it that the same compiler you used to install SIESTA? It doesn't have to be, but if SIESTA compiles you should use what works there. Your FC line should be pointing to f90, or whatever fortran compiler (pg90, ifort, ifc, gfortran, etc.) you used in the SIESTA install. Check your Obj/ folder in the SIESTA source tree for the arch.make.
I could not reproduce your error when I used my own f77 compiler or f90 or gfortran so that's strange. Try using gfortran, perhaps? You should not have to edit the source code to get this utility to work. Just try a different compiler until you can get it to work. My gfortran is: GNU Fortran 95 (GCC) 4.1.2 20070115 and it compiles without error. Swarm is Open-MP enabled, so you could compile it in parallel as well. On Thu, Jun 11, 2015 at 6:29 AM, Ludwig, Stephan < [email protected]> wrote: > > Hello Abraham, > > > thank you for your suggestion. Unfortunately I have proplems to install > the optimization programm. > > typing 'make swarm simplex' I receive the error message > > i1@QuanQal2:/home/siesta-3.2-pl-5/Util/Optimizer$ sudo make swarm simplex > > f77 -c io.f > io_assign: > Error on line 2: syntax error > Warning on line 13: ignoring text after "end". > Warning on line 13: local variable out never used > Warning on line 13: local variable intend never used > Error on line 13: Declaration error for intend: adjustable dimension on > non-argument > /usr/bin/f77: aborting compilation > make: *** [io.o] Fehler 25 > > The io.f file looks like this > > > SUBROUTINE io_assign(lun) > integer intend(out) :: lun > logical used > integer iostat > c > c Looks for a free unit and assigns it to lun > c > do lun= 10, 99 > inquire(unit=lun, opened=used, iostat=iostat) > if (iostat .ne. 0) used = .true. > if (.not. used) return > enddo > end SUBROUTINE io_assign > > > > Can you help me to find the syntax error? > > > Best regards > > Stephan > > > > > -----Original message----- > *From:* Abraham Hmiel <[email protected]> > *Sent:* Wednesday 10th June 2015 17:42 > *To:* Siesta, Self-Consistent DFT LCAO program, http://www.uam.es/siesta < > [email protected]> > *Subject:* Re: [SIESTA-L] PAO-basis block for As > > Stephan, > > Creating a PAO basis set from scratch is a nontrivial exercise even for > experts. The way I usually use involves the simplex tetrahedral method, > which allows you to vary several parameters at a time and find their > optimized values. There is an implementation included with SIESTA. > > Look in the Util/Optimizer folder of the siesta source tree. There are > some examples there which will allow you to use the simplex method code in > a basic manner for a water molecule to optimize basis set parameters. Take > a look at the files and study them and run the example calculations. > They're pretty quick. From there, it's not a big stretch to apply the > principles to arsenic. You will still have to figure out the format of the > PAO basis block corresponding to your pseudopotential and desired size of > the basis set. You might try using As2S3 or solid arsenic as input, and > optimize for bond length and total energy minimization. I'd offer that you > might learn more doing it this way. > > Kind regards, > > On Wed, Jun 10, 2015 at 7:27 AM, Nick Papior Andersen < > [email protected]> wrote: > >> I would suggest you ask your supervisor/professor (I suspect you have >> one?). >> >> 2015-06-10 16:24 GMT+02:00 Ludwig, Stephan < >> [email protected]>: >> >>> Hello everybody, >>> >>> >>> I'm an absolute beginner in DFT calculations (I'm a Master student) and >>> now I have to cope with a material containing Arsenid. >>> >>> The pseudopotential of As in the database has a semicore state so I have >>> to implement the PAO-basis block in the fdf-file. >>> >>> >>> I have no idea how to choose the parameters of this block to obtain >>> reasonable results. >>> >>> I was able to start siesta by using a PAO-basis block which was >>> originally designed for iron: >>> >>> >>> %block PAO.Basis # Define Basis set >>> As 1 # Species label, number of >>> l-shells >>> n=3 2 2 # n, l, Nzeta >>> 6.000 0.000 >>> 1.000 1.000 >>> %endblock PAO.Basis >>> >>> >>> >>> >>> Can anybody please deliver a suitable PAO-basis and a pseudopotential >>> for As. >>> >>> Or can anybody help me to find a suitable PAO-basis for the >>> pseudopotential I use. You can find it in the attachment >>> >>> It would be a great help and I would learn a lot through it >>> >>> >>> Thanks and regards >>> >>> >>> Stephan Ludwig >>> >>> >>> >>> >> >> >> -- >> Kind regards Nick >> > > > > -- > *Abraham Hmiel, Ph. D.* > http://abehmiel.net/about > > -----Original message----- > *From:* Abraham Hmiel <[email protected]> > *Sent:* Wednesday 10th June 2015 17:42 > *To:* Siesta, Self-Consistent DFT LCAO program, http://www.uam.es/siesta < > [email protected]> > *Subject:* Re: [SIESTA-L] PAO-basis block for As > > Stephan, > > Creating a PAO basis set from scratch is a nontrivial exercise even for > experts. The way I usually use involves the simplex tetrahedral method, > which allows you to vary several parameters at a time and find their > optimized values. There is an implementation included with SIESTA. > > Look in the Util/Optimizer folder of the siesta source tree. There are > some examples there which will allow you to use the simplex method code in > a basic manner for a water molecule to optimize basis set parameters. Take > a look at the files and study them and run the example calculations. > They're pretty quick. From there, it's not a big stretch to apply the > principles to arsenic. You will still have to figure out the format of the > PAO basis block corresponding to your pseudopotential and desired size of > the basis set. You might try using As2S3 or solid arsenic as input, and > optimize for bond length and total energy minimization. I'd offer that you > might learn more doing it this way. > > Kind regards, > > On Wed, Jun 10, 2015 at 7:27 AM, Nick Papior Andersen < > [email protected]> wrote: > >> I would suggest you ask your supervisor/professor (I suspect you have >> one?). >> >> 2015-06-10 16:24 GMT+02:00 Ludwig, Stephan < >> [email protected]>: >> >>> Hello everybody, >>> >>> >>> I'm an absolute beginner in DFT calculations (I'm a Master student) and >>> now I have to cope with a material containing Arsenid. >>> >>> The pseudopotential of As in the database has a semicore state so I have >>> to implement the PAO-basis block in the fdf-file. >>> >>> >>> I have no idea how to choose the parameters of this block to obtain >>> reasonable results. >>> >>> I was able to start siesta by using a PAO-basis block which was >>> originally designed for iron: >>> >>> >>> %block PAO.Basis # Define Basis set >>> As 1 # Species label, number of >>> l-shells >>> n=3 2 2 # n, l, Nzeta >>> 6.000 0.000 >>> 1.000 1.000 >>> %endblock PAO.Basis >>> >>> >>> >>> >>> Can anybody please deliver a suitable PAO-basis and a pseudopotential >>> for As. >>> >>> Or can anybody help me to find a suitable PAO-basis for the >>> pseudopotential I use. You can find it in the attachment >>> >>> It would be a great help and I would learn a lot through it >>> >>> >>> Thanks and regards >>> >>> >>> Stephan Ludwig >>> >>> >>> >>> >> >> >> -- >> Kind regards Nick >> > > > > -- > *Abraham Hmiel, Ph. D.* > http://abehmiel.net/about > > -- *Abraham Hmiel, Ph. D.* http://abehmiel.net/about
