supercell_dim

Syntax

supercell_dim = 12 12 16
Meaning

Number of primitive-cell repeats used to build the MD supercell. This should match the actual MD trajectory cell, because pySED uses it to decide which q-points are commensurate with the simulation.

Default

1 1 1.

Notes

This is one of the most important parameters in input_SED.in because it determines the repetition matrix used for commensurate q-point selection. pySED does not use a separate num_qpoints parameter; the available q-points are determined by supercell_dim, the primitive cell, the trajectory cell, and q_path.

For a simple diagonal supercell with repetition matrix \(P=\mathrm{diag}(N_x,N_y,N_z)\), a path from Gamma to the zone boundary along one reduced reciprocal direction, for example q_path = 0 0 0  0 0 0.5, gives allowed points \(0, 1/N_z, 2/N_z, \ldots, \lfloor N_z/2 \rfloor/N_z\). The number of q-points is therefore \(\lfloor N_z/2 \rfloor + 1\), or \(N_z/2 + 1\) when \(N_z\) is even.

A non-orthogonal real-space cell does not by itself change this counting if the supercell is still a diagonal repetition of the primitive lattice in reduced coordinates. What matters is the integer repetition matrix \(P\) satisfying supercell = P @ primitive.

For a non-diagonal or transformed supercell, the number of q-points depends on the selected path direction. pySED keeps only fractional positions \(f\) on the line \(q(f)=q_{\mathrm{start}}+f(q_{\mathrm{end}}-q_{\mathrm{start}})\) that satisfy \(q(f)P^T \in \mathbb{Z}^3\). In this case, inspect the printed Number of q-points generated message or the output .Qpts file. Increase the supercell repeats in the lattice directions that project onto the desired q-path to improve resolution. If the calculation can still fit in memory, use as large a supercell as practical; a larger supercell gives denser q-point sampling.

Back to Parameter Index