Dear Xavier,

Thanks for your quick answer. The reason behind this question is that I'm trying to fit the spin density extracted from the Kwant spin density operator with a formula derived from the spin diffusion equations (to extract e.g. the spin diffusion length or spin Hall angle). In this reference (https://arxiv.org/abs/1908.11354) they seem to complement the information in the Kwant documentation, and effectively, the units of the (spin) density are 1/energy. However, the solution of the spin density in the diffusion equations is 1/length^2 (in 2 dimensions). So a priori we did not know how to relate them, so we thought of dividing the output of the spin density operator with the output of the current density operator to get rid of the energy terms in the Kwant output. But if the current is unitless then I guess this is not useful... Therefore, let me ask you another question, how can I relate the spin density extracted from kwant with the one derived from the diffusion equations?

Thanks a lot again.

Best regards,

Marc