消除dicekriging的不断趋势

Question

我想通过公式inpu

中的“-1”来消除常量趋势

design = sin(rnorm(5))
obs = rnorm(5)
model <- km(formula = ~ -1 + sin(x), design = data.frame(x = design),
            response = obs, covtype = "gauss")
model  # shows the model

输出

Call:
km(formula = ~-1 + sin(x), design = data.frame(x = design), response = obs, 
    covtype = "gauss")

Trend  coeff.:
                Estimate
 (Intercept)    -0.1000
      sin(x)     0.8544

Covar. type  : gauss 
Covar. coeff.:
           Estimate
theta(x)     0.0000

Variance estimate: 0.7597348

恒定趋势仍然存在（通过截距估计显示）。有没有办法消除DiceKriging中的趋势？

非常感谢你的帮助。

Answer 1

我有过DiceKriging的经验

修改您的代码

design = sin(rnorm(5))
obs = rnorm(5)
model <- km(formula = NULL, design = data.frame(x = design),
    response = obs, covtype = "gauss")
model  # shows the model