Hi everyone,
I use the MNP package to fit my data. I can interpret the coefficients and
covariance. I cannot put them into the original model which is
Wi=Sum(bXij) over j,
Yi(Wi)=0 if max(Wi)<0, j if max(Wi)=Wij>0
j=3 here, 3 choice alternatives, Clio, Punto, Polo
I don't know how to put the coefficients and covariances into W, to
construct W for each j.
The coefficients and covariances are below.
I f anyone help me on this, I would appreciate.
ozlem.
Call:
mnp(formula = choice ~ 1, data = rowdata3, choiceX = list(Punto PricePunto,
Clio = PriceClio, Polo = PricePolo), cXnames = "price", base =
"Polo",
n.draws = 100, verbose = TRUE)
Coefficients:
mean std.dev. 2.5% 97.5%
(Intercept):Clio -1.4559 0.6889 -2.7017 -0.387
(Intercept):Punto -0.1989 0.7180 -2.2192 0.893
price 10.0865 5.1486 0.8549 23.030
Covariances:
mean std.dev. 2.5% 97.5%
Clio:Clio 1.0000 0.0000 1.0000 1.000
Clio:Punto -0.9458 0.7644 -2.1816 0.449
Punto:Punto 2.5980 2.5312 0.4339 9.281
Base category: Polo
Number of alternatives: 3
Number of observations: 12
Number of estimated parameters: 5
Number of stored MCMC draws: 100
[[alternative HTML version deleted]]