@@ -36,7 +36,7 @@ With a mixture of the following properties:

<sup>1</sup> : Stationary results checks are not entirely conclusive for $`t \to \infty`$ due to the fact that the simulations haven't been performed for times long enough so that the stationary state is reached. Here are the currents for the lead number $0$ for the system simulated by `2D-square-lattice/disorder_mag_field.py`

Update [14/07/2021]: the oscillations of the computed quantities at long times are due to the refine algorithm. It refines for the shape of the integrand at the beginning of the simulation, for small times. The issue is that the integrand's shape evolves with time but refining does not evolve and ends up being not adapted anymore, adding even more intervals for bigger times may help, but the maximum number of intervals would probably be quickly reached. For now, it is recommended to use an equidistant static initial split instead of an automatic refine.

Update [14/07/2021]: the oscillations of the computed quantities at long times are due to the refine algorithm. It refines for the shape of the integrand at the beginning of the simulation, for small times. The issue is that the integrand's shape evolves with time but refining does not evolve and ends up being not adapted anymore, adding even more intervals for bigger times may help, but the maximum number of intervals would probably be quickly reached. For now, it is recommended to use an equidistant static initial split instead of an automatic refine. The static equidistant refine has not been used while performing these validations.

![alt text](2D-square-lattice/data-analysis/disorder_mag_field_twice_refine_energy_refine_op_absorbing_boundary_lead_0_plot.png"Lead 0 Currents of 2D-square-lattice/disorder_mag_field.py")