Dear Rahman Thank you for your kind reply, I will rerun the calculations with your suggestions, and verify the relationship between the parameters defined in the ProjectedDensityOfStates block and the DOS results.There is another question puzzled me these days. I get the band structure of the (primitive) unit cell (actually is LiF), and the unit of high-symmetry points defined in the Bandlines block is given by ReciprocalLatticeVectors option. When I am trying to do the same calculation with a supercell, can I use the previous Bandlines block (not changed) in the new supercell calculations. Because I think in the supercell calculations, the LatticeVectors is expanded several times, maybe I should find the new locations of the X, W, K and Gamma points inside the new BZ.RegardsFei Mao USC
在2015-08-12，Gul Rahman <[email protected]> 写道:-----原始邮件----- 发件人: Gul Rahman <[email protected]> 发送时间: 2015年8月12日 星期三 收件人: "[email protected]" <[email protected]> 主题: Re: [SIESTA-L] The problem of DOS and PDOS calculation Dear Fei, Good that you have solved your problem. Well, the peak width (an energy) is for broadening the eigenvalues. You can choose any number and plot the DOS to see how it looks like. It is for broadening, so it will affect the shape of your DOS. You can play with it and plot your DOS for different values of broadening for the same k-points. The number of points are just the number (number of points in the energy window) to connect two points in numerical integration. In the language of numerical methods, if you use large number of points then your function (in this case your DOS) will be like a curve, and if you use very small number then it will look like a line. We do the same when we plot the band structure. You can also play with it. Regards, Gul Dr. Gul Rahman Assistant Professor, Department of Physics, Quaid-i-Azam University, Islamabad, Pakistan http://www.qau.edu.pk/profile.php?id=818020 On Tuesday, 11 August 2015, 7:16, 毛飞 <[email protected]> wrote: >Dear Rahman Thank you for you kind reply, and your suggestion works! Can you explain to me why, is there any relationship or requirement for choosing the peak width and the number of points in the energy window in the block definition. Regards Fei 在2015-08-11，Gul Rahman <[email protected]> 写道:-----原始邮件----- 发件人: Gul Rahman <[email protected]> 发送时间: 2015年8月11日 星期二 收件人: "[email protected]" <[email protected]> 主题: Re: [SIESTA-L] The problem of DOS and PDOS calculation Try something like this %block ProjectedDensityOfStates -25.00 20.00 0.100 1500 eV %endblock ProjectedDensityOfStates Dr. Gul Rahman Assistant Professor, Department of Physics, Quaid-i-Azam University, Islamabad, Pakistan http://www.qau.edu.pk/profile.php?id=818020 On Monday, 10 August 2015, 18:38, 毛飞 <[email protected]> wrote: Dear Siesta Developers and Users I am trying to calculate DOS and PDOS. To calculate both I add a block label in the input file, such us:%block ProjectedDensityOfSTates -10.0 10.0 0.05 700 eV%endblock ProjectedDensityOfStatesI obtain two output files SystemLabel.DOS and SystemLabel.PDOS. However, the *.DOS is totally empty, and the content in the *.PDOS is, <pdos> <nspin>1</nspin> <norbitals> 18</norbitals> <energy_values units="eV"> </energy_values> It is strange, I cannot understand these results, and I have searched the mail archive of Siesta, but I don’t find the explanations to this problem. Dose anybody encounter the problem before? I am using the version of siesta-3.2-pl-5. Regards Fei Mao USC
