Carefully read the description of the package... it uses a particular variety
of splines. You should go read the paper they reference.
On May 12, 2020 4:37:32 PM PDT, Yousri Fanous wrote:
>Hello
>
>I created a MARS model from earth package.for one predictor EngDispl vs
>FE.
>These are the result
Hello
I created a MARS model from earth package.for one predictor EngDispl vs FE.
These are the resulting coefficients:
summary(marsFit)$coefficients
y
(Intercept) 18.049001
h(EngDispl-4.3) -4.378082
h(4.3-EngDispl) 10.925240
h(EngDispl-2.3) 5.942387
h(EngDispl-3.5)
2 matches
Mail list logo
