Hello Paul
No matter what the cause for periodicity you might model that by cos and sin. However I like to bring to your attention that this is a better form
Exp[d((x-a)^2+(x-b)^2)]*Cos[e(x-a)]*Cos[e(x-b)]
the more variable you add, the more the terms in the norm in the exp and more Cos terms.
The above gives flexible modelling with more control.
No matter what the cause for periodicity you might model that by cos and sin. However I like to bring to your attention that this is a better form
Exp[d((x-a)^2+(x-b)^2)]*Cos[e(x-a)]*Cos[e(x-b)]
the more variable you add, the more the terms in the norm in the exp and more Cos terms.
The above gives flexible modelling with more control.
Version 2.1.8p2
