Data of all figures and codes employed to generate the theoretical data of "Anomalous diffusion and localization in a disorder-free atomic mixture".
Abstract
Codes for the semiclassical simulations (solid lines in Figs. 2 and 3): 1) SemiClassicalSimulator.py2) SCSim_header.pyCode to evaluate the quantum diffusion constant (colormap in Fig 4d): DiffusionFromReturnProb.nbCodes for calculation of the mobility edge (white points in Fig. 4d): 1. grscalar.f: Computes eigenvalues Lambda of a large number of N x N random matrices exp(i*k*r_{mn})/k*r_{mn}, with r_{mn} being the distance between resonant point scatterers m and n located at random and independent positions inside a sphere of radius R. N = total number of scatterers, rho = N/V their number density, V = (4*pi/3)*R^3 = volume of the sphere. 2. percentile.f: Uses arrays of eigenvalues generated by grscalar.f to compute q-th percentiles of the probability distribution of Thouless conductance g (q = 0.01-0.1). 3. percentile.nb: Mathematica code to analyse percentiles and to locate the mobility edges by finding a crossing point of percentiles calculated for different N, as a function of frequency omega = omega_0 - (Gamma_0/2)*Re(Lambda). Details of methodology are described in details in the following references:1. Skipetrov, S.E. Finite-size scaling analysis of localization transition for scalar waves in a three-dimensional ensemble of resonant point scatterers. Phys. Rev. B 94, 064202 (2016).2. Skipetrov, S.E. & Sokolov, I.M. Ioffe-Regel criterion for Anderson localization in the model of resonant scatterers. Phys. Rev. B 98, 064207 (2018).
