Re: [SIESTA-L] << Increasing RHO mesh >>
I have looked into it. And as far as I can see, it just takes a really long time... Like, many hours (at least). I can't really say how long it will take for this system. Denchar is sadly not parallel and could use an overhaul on that part. Den fre. 16. apr. 2021 kl. 22.02 skrev I. Camps : > Hi Nick, > > Here are two links with all the inputs for DENCHAR: > Link 1: > https://lamodel.unifal-mg.edu.br/downloads/Camps_DencharInput.tar.gz > Link 2: > https://drive.google.com/file/d/1GyjLmpOEZ553-A7R8r9oXqwlzx9NcQGb/view?usp=sharing > > Also, I compiled DENCHAR with *-debug all -g* options (I am using Intel > suite), but no new message appear. > > Thank you for your disponibility. > > []'s, > > Camps > > > On Wed, Apr 14, 2021 at 5:01 PM Nick Papior wrote: > >> Could you perhaps create a complete tar.gz so that I can test. >> Otherwise, you may recompile denchar with full debug options, that should >> give more details when it crashes. >> >> Den fre. 9. apr. 2021 kl. 22.07 skrev I. Camps : >> >>> Hello Nick, >>> >>> My input file is below. I duplicate the grid size. >>> >>> Also, I save the output from DENCHAR to a file and keep monitoring the >>> CPU and memory usage. >>> >>> In the output there is no indication of error and the memory usage keep >>> above 4GB with a 49% use (my note have 8GB). >>> >>> ### INPUT >>> SystemLabel BN-Ni-p4-dCRITIC2 >>> NumberOfSpecies3 >>> %block ChemicalSpeciesLabel >>> 1 5 B >>> 2 7 N >>> 3 28 Ni >>> %endblock ChemicalSpeciesLabel >>> Denchar.TypeOfRun 3D >>> Denchar.PlotCharge T >>> Denchar.CoorUnits Ang # Format for coordinate of the >>> points >>># Bohr >>># Ang >>> Denchar.DensityUnits Ele/Ang**3 # Units of Charge Density >>># Ele/bohr**3 >>># Ele/Ang**3 >>># Ele/UnitCell >>> Denchar.NumberPointsX 480 >>> Denchar.NumberPointsY 480 >>> Denchar.NumberPointsZ 360 >>> >>> Denchar.MinX 0 Ang >>> Denchar.MaxX 17.833623 Ang >>> Denchar.MinY 0 Ang >>> Denchar.MaxY 17.833623 Ang >>> Denchar.MinZ 0 Ang >>> Denchar.MaxZ 12.171076 Ang >>> ### >>> >>> ### OUTPUT >>> ... >>> ... >>> 115 23.089612.3513 2.6562 >>>116 23.041412.373710.7105 >>>117 23.060412.365518.7637 >>>118 24.088814.5243 1.3261 >>>119 24.059614.5373 9.3802 >>>120 24.062314.539317.4349 >>>121 14.8510 5.857515.3723 >>> >>>Generating Charge Density values on the grid now >>>Please, wait >>> ### >>> >>> >>> Regards, >>> >>> Camps >>> >>> >>> On Wed, Apr 7, 2021 at 5:17 PM Nick Papior wrote: >>> Hi, Hmm.. That depends on the options you use. It could be that you run out of memory? Perhaps some more information about what was before this would be useful? I.e. what does denchar print before writing "Killed"? Den tor. 1. apr. 2021 kl. 22.02 skrev I. Camps : > Thank you very much Nick! > > I am using the second option... > > The original mesh is 240x240x180. I tried to double it, but DENCHAR > just stopped with a simple text: "Killed", without any information about. > > Do you have a recommendation about the best form to select the grid? > > []'s, > > Camps > > > On Mon, Mar 29, 2021 at 5:02 PM Nick Papior > wrote: > >> Hi, >> >> You can use sisl to interpolate and write out the a cube file. >> >> For instance, >> >> sgrid siesta.RHO --interp 300 400 400 --out rho.cube >> >> will interpolate the RHO grid to a size of 300, 400, 400, x, y, z >> directions. Please refer to sgrid --help for more information. For >> instance >> the interpolation method may be quite important in your case. >> >> Alternatively you can use denchar to replot the density matrix files >> using a denser grid. This is of course more precise since it isn't >> interpolation but using the same DM to recreate the full grid on a >> different mesh. >> >> Good luck! >> >> Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : >> >>> Dear SIESTers, >>> >>> I am doing a critical point study using the package CRITIC2, but for >>> SIESTA, it only accepts the RHO density in cube format, so its results >>> are >>> dependent on the quality of the used grid. >>> >>> Is it possible to increase the mesh for the grid where RHO is >>> calculated? >>> >>> []'s, >>> >>> Camps >>> >>> -- >>> SIESTA is supported by the Spanish Research Agency (AEI) and
Re: [SIESTA-L] << Increasing RHO mesh >>
Hi Nick, Here are two links with all the inputs for DENCHAR: Link 1: https://lamodel.unifal-mg.edu.br/downloads/Camps_DencharInput.tar.gz Link 2: https://drive.google.com/file/d/1GyjLmpOEZ553-A7R8r9oXqwlzx9NcQGb/view?usp=sharing Also, I compiled DENCHAR with *-debug all -g* options (I am using Intel suite), but no new message appear. Thank you for your disponibility. []'s, Camps On Wed, Apr 14, 2021 at 5:01 PM Nick Papior wrote: > Could you perhaps create a complete tar.gz so that I can test. > Otherwise, you may recompile denchar with full debug options, that should > give more details when it crashes. > > Den fre. 9. apr. 2021 kl. 22.07 skrev I. Camps : > >> Hello Nick, >> >> My input file is below. I duplicate the grid size. >> >> Also, I save the output from DENCHAR to a file and keep monitoring the >> CPU and memory usage. >> >> In the output there is no indication of error and the memory usage keep >> above 4GB with a 49% use (my note have 8GB). >> >> ### INPUT >> SystemLabel BN-Ni-p4-dCRITIC2 >> NumberOfSpecies3 >> %block ChemicalSpeciesLabel >> 1 5 B >> 2 7 N >> 3 28 Ni >> %endblock ChemicalSpeciesLabel >> Denchar.TypeOfRun 3D >> Denchar.PlotCharge T >> Denchar.CoorUnits Ang # Format for coordinate of the >> points >># Bohr >># Ang >> Denchar.DensityUnits Ele/Ang**3 # Units of Charge Density >># Ele/bohr**3 >># Ele/Ang**3 >># Ele/UnitCell >> Denchar.NumberPointsX 480 >> Denchar.NumberPointsY 480 >> Denchar.NumberPointsZ 360 >> >> Denchar.MinX 0 Ang >> Denchar.MaxX 17.833623 Ang >> Denchar.MinY 0 Ang >> Denchar.MaxY 17.833623 Ang >> Denchar.MinZ 0 Ang >> Denchar.MaxZ 12.171076 Ang >> ### >> >> ### OUTPUT >> ... >> ... >> 115 23.089612.3513 2.6562 >>116 23.041412.373710.7105 >>117 23.060412.365518.7637 >>118 24.088814.5243 1.3261 >>119 24.059614.5373 9.3802 >>120 24.062314.539317.4349 >>121 14.8510 5.857515.3723 >> >>Generating Charge Density values on the grid now >>Please, wait >> ### >> >> >> Regards, >> >> Camps >> >> >> On Wed, Apr 7, 2021 at 5:17 PM Nick Papior wrote: >> >>> Hi, >>> >>> Hmm.. That depends on the options you use. It could be that you run out >>> of memory? >>> Perhaps some more information about what was before this would be >>> useful? I.e. what does denchar print before writing "Killed"? >>> >>> Den tor. 1. apr. 2021 kl. 22.02 skrev I. Camps : >>> Thank you very much Nick! I am using the second option... The original mesh is 240x240x180. I tried to double it, but DENCHAR just stopped with a simple text: "Killed", without any information about. Do you have a recommendation about the best form to select the grid? []'s, Camps On Mon, Mar 29, 2021 at 5:02 PM Nick Papior wrote: > Hi, > > You can use sisl to interpolate and write out the a cube file. > > For instance, > > sgrid siesta.RHO --interp 300 400 400 --out rho.cube > > will interpolate the RHO grid to a size of 300, 400, 400, x, y, z > directions. Please refer to sgrid --help for more information. For > instance > the interpolation method may be quite important in your case. > > Alternatively you can use denchar to replot the density matrix files > using a denser grid. This is of course more precise since it isn't > interpolation but using the same DM to recreate the full grid on a > different mesh. > > Good luck! > > Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : > >> Dear SIESTers, >> >> I am doing a critical point study using the package CRITIC2, but for >> SIESTA, it only accepts the RHO density in cube format, so its results >> are >> dependent on the quality of the used grid. >> >> Is it possible to increase the mesh for the grid where RHO is >> calculated? >> >> []'s, >> >> Camps >> >> -- >> SIESTA is supported by the Spanish Research Agency (AEI) and by the >> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >> > > > -- > Kind regards Nick > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence
Re: [SIESTA-L] << Increasing RHO mesh >>
Could you perhaps create a complete tar.gz so that I can test. Otherwise, you may recompile denchar with full debug options, that should give more details when it crashes. Den fre. 9. apr. 2021 kl. 22.07 skrev I. Camps : > Hello Nick, > > My input file is below. I duplicate the grid size. > > Also, I save the output from DENCHAR to a file and keep monitoring the CPU > and memory usage. > > In the output there is no indication of error and the memory usage keep > above 4GB with a 49% use (my note have 8GB). > > ### INPUT > SystemLabel BN-Ni-p4-dCRITIC2 > NumberOfSpecies3 > %block ChemicalSpeciesLabel > 1 5 B > 2 7 N > 3 28 Ni > %endblock ChemicalSpeciesLabel > Denchar.TypeOfRun 3D > Denchar.PlotCharge T > Denchar.CoorUnits Ang # Format for coordinate of the points ># Bohr ># Ang > Denchar.DensityUnits Ele/Ang**3 # Units of Charge Density ># Ele/bohr**3 ># Ele/Ang**3 ># Ele/UnitCell > Denchar.NumberPointsX 480 > Denchar.NumberPointsY 480 > Denchar.NumberPointsZ 360 > > Denchar.MinX 0 Ang > Denchar.MaxX 17.833623 Ang > Denchar.MinY 0 Ang > Denchar.MaxY 17.833623 Ang > Denchar.MinZ 0 Ang > Denchar.MaxZ 12.171076 Ang > ### > > ### OUTPUT > ... > ... > 115 23.089612.3513 2.6562 >116 23.041412.373710.7105 >117 23.060412.365518.7637 >118 24.088814.5243 1.3261 >119 24.059614.5373 9.3802 >120 24.062314.539317.4349 >121 14.8510 5.857515.3723 > >Generating Charge Density values on the grid now >Please, wait > ### > > > Regards, > > Camps > > > On Wed, Apr 7, 2021 at 5:17 PM Nick Papior wrote: > >> Hi, >> >> Hmm.. That depends on the options you use. It could be that you run out >> of memory? >> Perhaps some more information about what was before this would be useful? >> I.e. what does denchar print before writing "Killed"? >> >> Den tor. 1. apr. 2021 kl. 22.02 skrev I. Camps : >> >>> Thank you very much Nick! >>> >>> I am using the second option... >>> >>> The original mesh is 240x240x180. I tried to double it, but DENCHAR just >>> stopped with a simple text: "Killed", without any information about. >>> >>> Do you have a recommendation about the best form to select the grid? >>> >>> []'s, >>> >>> Camps >>> >>> >>> On Mon, Mar 29, 2021 at 5:02 PM Nick Papior >>> wrote: >>> Hi, You can use sisl to interpolate and write out the a cube file. For instance, sgrid siesta.RHO --interp 300 400 400 --out rho.cube will interpolate the RHO grid to a size of 300, 400, 400, x, y, z directions. Please refer to sgrid --help for more information. For instance the interpolation method may be quite important in your case. Alternatively you can use denchar to replot the density matrix files using a denser grid. This is of course more precise since it isn't interpolation but using the same DM to recreate the full grid on a different mesh. Good luck! Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : > Dear SIESTers, > > I am doing a critical point study using the package CRITIC2, but for > SIESTA, it only accepts the RHO density in cube format, so its results are > dependent on the quality of the used grid. > > Is it possible to increase the mesh for the grid where RHO is > calculated? > > []'s, > > Camps > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- Kind regards Nick -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >>> >>> -- >>> SIESTA is supported by the Spanish Research Agency (AEI) and by the >>> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >>> >> >> >> -- >> Kind regards Nick >> >> -- >> SIESTA is supported by the Spanish Research Agency (AEI) and by the >> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >> > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- Kind regards Nick -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
Re: [SIESTA-L] << Increasing RHO mesh >>
Hello Nick, My input file is below. I duplicate the grid size. Also, I save the output from DENCHAR to a file and keep monitoring the CPU and memory usage. In the output there is no indication of error and the memory usage keep above 4GB with a 49% use (my note have 8GB). ### INPUT SystemLabel BN-Ni-p4-dCRITIC2 NumberOfSpecies3 %block ChemicalSpeciesLabel 1 5 B 2 7 N 3 28 Ni %endblock ChemicalSpeciesLabel Denchar.TypeOfRun 3D Denchar.PlotCharge T Denchar.CoorUnits Ang # Format for coordinate of the points # Bohr # Ang Denchar.DensityUnits Ele/Ang**3 # Units of Charge Density # Ele/bohr**3 # Ele/Ang**3 # Ele/UnitCell Denchar.NumberPointsX 480 Denchar.NumberPointsY 480 Denchar.NumberPointsZ 360 Denchar.MinX 0 Ang Denchar.MaxX 17.833623 Ang Denchar.MinY 0 Ang Denchar.MaxY 17.833623 Ang Denchar.MinZ 0 Ang Denchar.MaxZ 12.171076 Ang ### ### OUTPUT ... ... 115 23.089612.3513 2.6562 116 23.041412.373710.7105 117 23.060412.365518.7637 118 24.088814.5243 1.3261 119 24.059614.5373 9.3802 120 24.062314.539317.4349 121 14.8510 5.857515.3723 Generating Charge Density values on the grid now Please, wait ### Regards, Camps On Wed, Apr 7, 2021 at 5:17 PM Nick Papior wrote: > Hi, > > Hmm.. That depends on the options you use. It could be that you run out of > memory? > Perhaps some more information about what was before this would be useful? > I.e. what does denchar print before writing "Killed"? > > Den tor. 1. apr. 2021 kl. 22.02 skrev I. Camps : > >> Thank you very much Nick! >> >> I am using the second option... >> >> The original mesh is 240x240x180. I tried to double it, but DENCHAR just >> stopped with a simple text: "Killed", without any information about. >> >> Do you have a recommendation about the best form to select the grid? >> >> []'s, >> >> Camps >> >> >> On Mon, Mar 29, 2021 at 5:02 PM Nick Papior wrote: >> >>> Hi, >>> >>> You can use sisl to interpolate and write out the a cube file. >>> >>> For instance, >>> >>> sgrid siesta.RHO --interp 300 400 400 --out rho.cube >>> >>> will interpolate the RHO grid to a size of 300, 400, 400, x, y, z >>> directions. Please refer to sgrid --help for more information. For instance >>> the interpolation method may be quite important in your case. >>> >>> Alternatively you can use denchar to replot the density matrix files >>> using a denser grid. This is of course more precise since it isn't >>> interpolation but using the same DM to recreate the full grid on a >>> different mesh. >>> >>> Good luck! >>> >>> Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : >>> Dear SIESTers, I am doing a critical point study using the package CRITIC2, but for SIESTA, it only accepts the RHO density in cube format, so its results are dependent on the quality of the used grid. Is it possible to increase the mesh for the grid where RHO is calculated? []'s, Camps -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >>> >>> >>> -- >>> Kind regards Nick >>> >>> -- >>> SIESTA is supported by the Spanish Research Agency (AEI) and by the >>> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >>> >> >> -- >> SIESTA is supported by the Spanish Research Agency (AEI) and by the >> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >> > > > -- > Kind regards Nick > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
Re: [SIESTA-L] << Increasing RHO mesh >>
Hi, Hmm.. That depends on the options you use. It could be that you run out of memory? Perhaps some more information about what was before this would be useful? I.e. what does denchar print before writing "Killed"? Den tor. 1. apr. 2021 kl. 22.02 skrev I. Camps : > Thank you very much Nick! > > I am using the second option... > > The original mesh is 240x240x180. I tried to double it, but DENCHAR just > stopped with a simple text: "Killed", without any information about. > > Do you have a recommendation about the best form to select the grid? > > []'s, > > Camps > > > On Mon, Mar 29, 2021 at 5:02 PM Nick Papior wrote: > >> Hi, >> >> You can use sisl to interpolate and write out the a cube file. >> >> For instance, >> >> sgrid siesta.RHO --interp 300 400 400 --out rho.cube >> >> will interpolate the RHO grid to a size of 300, 400, 400, x, y, z >> directions. Please refer to sgrid --help for more information. For instance >> the interpolation method may be quite important in your case. >> >> Alternatively you can use denchar to replot the density matrix files >> using a denser grid. This is of course more precise since it isn't >> interpolation but using the same DM to recreate the full grid on a >> different mesh. >> >> Good luck! >> >> Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : >> >>> Dear SIESTers, >>> >>> I am doing a critical point study using the package CRITIC2, but for >>> SIESTA, it only accepts the RHO density in cube format, so its results are >>> dependent on the quality of the used grid. >>> >>> Is it possible to increase the mesh for the grid where RHO is calculated? >>> >>> []'s, >>> >>> Camps >>> >>> -- >>> SIESTA is supported by the Spanish Research Agency (AEI) and by the >>> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >>> >> >> >> -- >> Kind regards Nick >> >> -- >> SIESTA is supported by the Spanish Research Agency (AEI) and by the >> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >> > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- Kind regards Nick -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
Re: [SIESTA-L] << Increasing RHO mesh >>
Thank you very much Nick! I am using the second option... The original mesh is 240x240x180. I tried to double it, but DENCHAR just stopped with a simple text: "Killed", without any information about. Do you have a recommendation about the best form to select the grid? []'s, Camps On Mon, Mar 29, 2021 at 5:02 PM Nick Papior wrote: > Hi, > > You can use sisl to interpolate and write out the a cube file. > > For instance, > > sgrid siesta.RHO --interp 300 400 400 --out rho.cube > > will interpolate the RHO grid to a size of 300, 400, 400, x, y, z > directions. Please refer to sgrid --help for more information. For instance > the interpolation method may be quite important in your case. > > Alternatively you can use denchar to replot the density matrix files using > a denser grid. This is of course more precise since it isn't interpolation > but using the same DM to recreate the full grid on a different mesh. > > Good luck! > > Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : > >> Dear SIESTers, >> >> I am doing a critical point study using the package CRITIC2, but for >> SIESTA, it only accepts the RHO density in cube format, so its results are >> dependent on the quality of the used grid. >> >> Is it possible to increase the mesh for the grid where RHO is calculated? >> >> []'s, >> >> Camps >> >> -- >> SIESTA is supported by the Spanish Research Agency (AEI) and by the >> European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) >> > > > -- > Kind regards Nick > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
Re: [SIESTA-L] << Increasing RHO mesh >>
Hi, You can use sisl to interpolate and write out the a cube file. For instance, sgrid siesta.RHO --interp 300 400 400 --out rho.cube will interpolate the RHO grid to a size of 300, 400, 400, x, y, z directions. Please refer to sgrid --help for more information. For instance the interpolation method may be quite important in your case. Alternatively you can use denchar to replot the density matrix files using a denser grid. This is of course more precise since it isn't interpolation but using the same DM to recreate the full grid on a different mesh. Good luck! Den søn. 28. mar. 2021 kl. 23.01 skrev I. Camps : > Dear SIESTers, > > I am doing a critical point study using the package CRITIC2, but for > SIESTA, it only accepts the RHO density in cube format, so its results are > dependent on the quality of the used grid. > > Is it possible to increase the mesh for the grid where RHO is calculated? > > []'s, > > Camps > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > -- Kind regards Nick -- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)