A note on magnetic modelling with remanence