Estimate the restricted Nested Logit

Massimo

New Member
Hi everyone, I find it very difficult to estimate the model proposed in the post title; in particular:
I have done the following code

install.packages("mlogit")
install.packages("nlme")
library("mlogit")
library("nlme")

data("HC", package="mlogit")

HC1 <- mlogit.data(HC, varying = c(2:8, 10:16), choice = "depvar", shape = "wide")

cooling.modes <- index(HC1)$alt %in% c('gcc', 'ecc', 'erc', 'hpc') room.modes <- index(HC1)$alt %in% c('erc', 'er')
HC1$icca[!cooling.modes] <- 0 HC1$occa[!cooling.modes] <- 0
HC1$inc.cooling <- HC1$inc.room <- 0
HC1$inc.cooling[cooling.modes] <- HC1$income[cooling.modes]
HC1$inc.room[room.modes] <- HC1$income[room.modes]
HC1\$int.cooling <- as.numeric(cooling.modes)

#then estimating the following model:
nl <- mlogit(depvar ~ ich + och +icca + occa + inc.room + inc.cooling + int.cooling | 0 , HC1,
nests = list(cooling = c('gcc','ecc','erc','hpc'),
other = c('gc', 'ec', 'er')), un.nest.el = FALSE)

summary(nl):

#Coefficients :
# Estimate Std. Error t-value Pr(>|t|)
#ich -0.562283 0.146145 -3.8474 0.0001194 ***
#och -0.895493 0.271861 -3.2939 0.0009880 ***
#icca -0.267062 0.150310 -1.7767 0.0756103 .
#occa -1.338514 1.264215 -1.0588 0.2897042
#inc.room -0.381441 0.096658 -3.9463 7.937e-05 ***
#inc.cooling 0.259932 0.062085 4.1867 2.830e-05 ***
#int.cooling -4.821927 5.528796 -0.8721 0.3831277
#iv.cooling 0.611529 0.188736 3.2401 0.0011947 **
#iv.other 0.378394 0.133617 2.8319 0.0046270 **
#---
#Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

#Log-Likelihood: -178.04

my problem is how to impose a value on one of the iv, binding the model before the estimate (for example: iv.cooling=1), how can I do?