Dear de Vries, Your "nst*" values are so big that no enough samples are used to obtain the averaged one that can compare with the experimental data. Best regards, Chaofu Wu