Dear Kwant users, I am trying to obtain the local density of states of a topological (finite) system by using KPM approximation. My problem is that the gap is too small (as I am considering realistic parameters) and this results in more states than the topologically protected ones appearing for zero energy LDoS. If I control the energy resolution the memory consumption increases too much. So my question is if there is any parameter I can control to have a better energy resolution only near zero energy or maybe somewhere I can compensate the memory consumption (for example, losing resolution at the local distribution in order to obtain better energy resolution). Thanks in advance.  Antônio Lucas Rigotti Manesco PhD fellow  University of São Paulo, Brazil
Dear Antonio, KPM by itself doesn't allow to adjust the resolution locally. If you need detailed information about a small part of the spectrum, sparse diagonalization is a more appropriate tool. Best, Anton On Tue, Nov 28, 2017 at 1:58 PM, Antonio Lucas Rigotti Manesco <antoniolrm@usp.br> wrote:
Dear Kwant users,
I am trying to obtain the local density of states of a topological (finite) system by using KPM approximation. My problem is that the gap is too small (as I am considering realistic parameters) and this results in more states than the topologically protected ones appearing for zero energy LDoS. If I control the energy resolution the memory consumption increases too much.
So my question is if there is any parameter I can control to have a better energy resolution only near zero energy or maybe somewhere I can compensate the memory consumption (for example, losing resolution at the local distribution in order to obtain better energy resolution).
Thanks in advance.  Antônio Lucas Rigotti Manesco PhD fellow  University of São Paulo, Brazil
Dear Anton, Thanks for your response. I will do that. Best, 20171128 11:01 GMT02:00 Anton Akhmerov <anton.akhmerov+kd@gmail.com>:
Dear Antonio,
KPM by itself doesn't allow to adjust the resolution locally. If you need detailed information about a small part of the spectrum, sparse diagonalization is a more appropriate tool.
Best, Anton
On Tue, Nov 28, 2017 at 1:58 PM, Antonio Lucas Rigotti Manesco <antoniolrm@usp.br> wrote:
Dear Kwant users,
I am trying to obtain the local density of states of a topological (finite) system by using KPM approximation. My problem is that the gap is too small (as I am considering realistic parameters) and this results in more states than the topologically protected ones appearing for zero energy LDoS. If I control the energy resolution the memory consumption increases too much.
So my question is if there is any parameter I can control to have a better energy resolution only near zero energy or maybe somewhere I can compensate the memory consumption (for example, losing resolution at the local distribution in order to obtain better energy resolution).
Thanks in advance.  Antônio Lucas Rigotti Manesco PhD fellow  University of São Paulo, Brazil
 Antônio Lucas Rigotti Manesco PhD fellow  University of São Paulo, Brazil
participants (2)

Anton Akhmerov

Antonio Lucas Rigotti Manesco