In our published paper entitled, 'All-optical spin switching under different spin configurations' by G P Zhang and M Murakami 2019 J. Phys.: Condens. Matter 31 345802, figures 5(b)-(d) use zero exchange interactions. It came to our attention when we tried to simulate a ferrimagnet recently. The attached three figures, which replace those old figures, are now correct that all the data are computed with exchange interaction J = 0.1 eV/PLANCK CONSTANT OVER TWO PI2. The differences between the old and new figures are difficult to detect visually. When we check the data directly, we find a clear difference due to the exchange interaction. All the other figures are correct. This change does not affect the conclusion and finding.
