| Library: | gam |
| See also: | gamfit gamopt intest intestpl gintest gintestpl |
| Quantlet: | gamout | |
| Description: | auxiliary quantlet, creates a nice output for GAM. |
| Usage: | gamout(t,y,m,b,const{,opt}) | |
| Input: | ||
| t | n x p matrix, the predictor variables. | |
| y | n x 1 vector, the response variables. | |
| m | n x dd matrix, estimated functions. | |
| b | p x 1 vector, estimated coefficients. | |
| const | scalar, the estimated constant. | |
| opt | optional list, ouput of gamopt. The order of the elements of the list is not important, parameters not given are replaced by defaults. See gamopt for more details. | |
library("gam")
n=100
b=1|2
p=rows(b)
x=2.*uniform(n,p)-1
t=2.*uniform(n,1)-1
y=x*b+t^3+normal(n)./2
opt=gamopt("x",x,"pl",1)
{m,b,const}=backfit(t,y,1,0,"qua",opt)
gamout(t,y,m,b,const,opt)
Display with estimated functions, data points and display with coefficients for discrete variables.