标题: 伞形采样时集合变量平均值偏离约束值 [打印本页] 作者Author: ljh123 时间: 2025-10-5 15:56 标题: 伞形采样时集合变量平均值偏离约束值 我用伞形采样计算一个酸分子的酸度系数。集合变量是O-H的配位数。我把窗口按照0.125埃的间隔从0.1分到0.9。但是发现部分窗口集合变量的直方图平均值是偏离约束值的。例如我约束值为1.25时，平均值为1.1。力常数我设置为1500kj/A^2。请问下大家这种偏离是否可以忽略呢？还是说要加大力常数，让平均值严格在1.25附近？ 作者Author: student0618 时间: 2025-10-5 23:41
check histogram overlap. if the overlap is not good enough, you may want to e.g. increase the value of force constant/number of windows.作者Author: ljh123 时间: 2025-10-6 01:32
student0618 发表于 2025-10-5 23:41
check histogram overlap. if the overlap is not good enough, you may want to e.g. increase the value ...
I think there is a good overlap, as each window has a almost similar degree of offset. The overlap between windows occurs at about half of the peak height or more. Can I consider this result as usable?