Jiuning Hu wrote:

I am wondering if the output from the transmission matrix t(E) (used to calculate G(E)) of kwant is calculated at E = 0, i.e., what we have is actually t(E=0) and G(E=0). Although this looks like that the Fermi level is fixed at E = 0, the actual Fermi level (e.g., relative to charge neutral point) can be changed via the onsite potential.

This has been discussed before, try searching for “temperature” in the history of this mailing list [1]. You then should be able to find relevant threads like [2] or [3]. In a nutshell, Kwant’s transmission() gives the conductance at zero temperature in the regime where the left and right chemical potentials differ only infinitesimally. From this, it is possible to calculate an (approximate) finite-temperature, finite-voltage conductance through numerical integration. This is explained in the literature, for example in S. Datta’s book that we mention in the tutorial [4]. [1] https://kwant-project.org/community [2] https://www.mail-archive.com/kwant-discuss@kwant-project.org/msg00208.html [3] https://www.mail-archive.com/kwant-discuss@kwant-project.org/msg00909.html [4] https://kwant-project.org/doc/1/tutorial/introduction#quantum-transport