#### Apollo standard script #####
library(apollo) # Load apollo package
# Test treatment effect
database <- database_full %>%
filter(!is.na(Treatment_new)) %>%
mutate(Dummy_Treated = case_when(Treatment_new == 1|Treatment_new == 2 ~ 1, TRUE ~ 0),
Dummy_Vol_Treated = case_when(Treatment_new == 5 |Treatment_new == 4 ~ 1, TRUE ~ 0),
Dummy_no_info = case_when(Treatment_new == 3 ~ 1, TRUE~0))
table(database$Dummy_Treated)
table(database$Dummy_Vol_Treated)
table(database$Dummy_no_info)
#initialize model
apollo_initialise()
### Set core controls
apollo_control = list(
modelName = "MXL_wtp_Case_D_X",
modelDescr = "MXL wtp space Case D Interactions",
indivID ="id",
mixing = TRUE,
HB= FALSE,
nCores = n_cores,
outputDirectory = "Estimation_results/mxl"
)
##### Define model parameters depending on your attributes and model specification! ####
# set values to 0 for conditional logit model
apollo_beta=c(mu_natural = 15,
mu_walking = -1,
mu_rent = -2,
ASC_sq = 0,
mu_ASC_sq_treated = 0,
mu_ASC_sq_vol_treated = 0,
mu_ASC_sq_no_info = 0,
mu_ASC_NR = 0,
mu_ASC_Age = 0,
mu_ASC_Income = 0,
mu_rent_treated = 0,
mu_rent_vol_treated = 0,
mu_rent_no_info = 0,
mu_rent_NR = 0,
mu_rent_Age = 0,
mu_rent_Income = 0,
mu_nat_treated =0,
mu_nat_vol_treated = 0,
mu_nat_no_info = 0,
mu_nat_NR = 0,
mu_nat_Age = 0,
mu_nat_Income = 0,
mu_walking_treated =0,
mu_walking_vol_treated = 0,
mu_walking_no_info = 0,
mu_walking_NR = 0,
mu_walking_Age = 0,
mu_walking_Income = 0,
sig_natural = 15,
sig_walking = 2,
sig_rent = 2,
sig_ASC_sq = 2)
### specify parameters that should be kept fixed, here = none
apollo_fixed = c()
### Set parameters for generating draws, use 2000 sobol draws
apollo_draws = list(
interDrawsType = "sobol",
interNDraws = n_draws,
interUnifDraws = c(),
interNormDraws = c("draws_natural", "draws_walking", "draws_rent", "draws_asc"),
intraDrawsType = "halton",
intraNDraws = 0,
intraUnifDraws = c(),
intraNormDraws = c()
)
### Create random parameters, define distribution of the parameters
apollo_randCoeff = function(apollo_beta, apollo_inputs){
randcoeff = list()
randcoeff[["b_mu_natural"]] = mu_natural + sig_natural * draws_natural
randcoeff[["b_mu_walking"]] = mu_walking + sig_walking * draws_walking
randcoeff[["b_mu_rent"]] = -exp(mu_rent + sig_rent * draws_rent)
randcoeff[["b_ASC_sq"]] = ASC_sq + sig_ASC_sq * draws_asc
return(randcoeff)
}
### validate
apollo_inputs = apollo_validateInputs()
apollo_probabilities=function(apollo_beta, apollo_inputs, functionality="estimate"){
### Function initialisation: do not change the following three commands
### Attach inputs and detach after function exit
apollo_attach(apollo_beta, apollo_inputs)
on.exit(apollo_detach(apollo_beta, apollo_inputs))
### Create list of probabilities P
P = list()
#### List of utilities (later integrated in mnl_settings below) ####
# Define utility functions here:
V = list()
V[['alt1']] = -(b_mu_rent + mu_rent_treated *Dummy_Treated + mu_rent_vol_treated * Dummy_Vol_Treated + mu_rent_no_info * Dummy_no_info
+ mu_rent_NR * Z_Mean_NR + mu_rent_Age * Age_mean + mu_rent_Income * Income_mean)*
(b_mu_natural * Naturalness_1 + b_mu_walking * WalkingDistance_1 +
+ mu_nat_treated * Naturalness_1 *Dummy_Treated + mu_nat_no_info * Naturalness_1 * Dummy_no_info
+ mu_nat_vol_treated * Naturalness_1 * Dummy_Vol_Treated
+ mu_walking_treated * WalkingDistance_1 *Dummy_Treated + mu_walking_no_info * WalkingDistance_1 * Dummy_no_info
+ mu_walking_vol_treated * WalkingDistance_1 * Dummy_Vol_Treated
+ mu_nat_NR * Z_Mean_NR *Naturalness_1 + mu_nat_Age * Age_mean * Naturalness_1
+ mu_nat_Income * Income_mean * Naturalness_1 + mu_walking_NR * Z_Mean_NR * WalkingDistance_1
+ mu_walking_Age * Age_mean * WalkingDistance_1 + mu_walking_Income * Income_mean * WalkingDistance_1
- Rent_1)
V[['alt2']] = -(b_mu_rent + mu_rent_treated *Dummy_Treated + mu_rent_vol_treated * Dummy_Vol_Treated + mu_rent_no_info * Dummy_no_info
+ mu_rent_NR * Z_Mean_NR + mu_rent_Age * Age_mean + mu_rent_Income * Income_mean)*
(b_mu_natural * Naturalness_2 + b_mu_walking * WalkingDistance_2
+ mu_nat_treated * Naturalness_2 *Dummy_Treated + mu_nat_no_info * Naturalness_2 * Dummy_no_info
+ mu_nat_vol_treated * Naturalness_2 * Dummy_Vol_Treated
+ mu_walking_treated * WalkingDistance_2 *Dummy_Treated + mu_walking_no_info * WalkingDistance_2 * Dummy_no_info
+ mu_walking_vol_treated * WalkingDistance_2 * Dummy_Vol_Treated
+ mu_nat_NR * Z_Mean_NR *Naturalness_2 + mu_nat_Age * Age_mean * Naturalness_2
+ mu_nat_Income * Income_mean * Naturalness_2 + mu_walking_NR * Z_Mean_NR * WalkingDistance_2
+ mu_walking_Age * Age_mean * WalkingDistance_2 + mu_walking_Income * Income_mean * WalkingDistance_2
- Rent_2)
V[['alt3']] = -(b_mu_rent + mu_rent_treated *Dummy_Treated + mu_rent_vol_treated * Dummy_Vol_Treated + mu_rent_no_info * Dummy_no_info
+ mu_rent_NR * Z_Mean_NR + mu_rent_Age * Age_mean + mu_rent_Income * Income_mean)*
(b_ASC_sq + b_mu_natural * Naturalness_3 + b_mu_walking * WalkingDistance_3
+ mu_nat_treated * Naturalness_3 *Dummy_Treated + mu_nat_no_info * Naturalness_3 * Dummy_no_info
+ mu_nat_vol_treated * Naturalness_3 * Dummy_Vol_Treated
+ mu_walking_treated * WalkingDistance_3 *Dummy_Treated + mu_walking_no_info * WalkingDistance_3 * Dummy_no_info
+ mu_walking_vol_treated * WalkingDistance_3 * Dummy_Vol_Treated
+ mu_ASC_sq_treated * Dummy_Treated + mu_ASC_sq_vol_treated * Dummy_Vol_Treated
+ mu_ASC_sq_no_info * Dummy_no_info
+ mu_ASC_NR * Z_Mean_NR + mu_ASC_Age * Age_mean + mu_ASC_Income * Income_mean
+ mu_nat_NR * Z_Mean_NR *Naturalness_3 + mu_nat_Age * Age_mean * Naturalness_3
+ mu_nat_Income * Income_mean * Naturalness_3 + mu_walking_NR * Z_Mean_NR * WalkingDistance_3
+ mu_walking_Age * Age_mean * WalkingDistance_3 + mu_walking_Income * Income_mean * WalkingDistance_3
- Rent_3)
### Define settings for MNL model component
mnl_settings = list(
alternatives = c(alt1=1, alt2=2, alt3=3),
avail = 1, # all alternatives are available in every choice
choiceVar = choice,
V = V#, # tell function to use list vector defined above
)
### Compute probabilities using MNL model
P[['model']] = apollo_mnl(mnl_settings, functionality)
### Take product across observation for same individual
P = apollo_panelProd(P, apollo_inputs, functionality)
### Average across inter-individual draws - nur bei Mixed Logit!
P = apollo_avgInterDraws(P, apollo_inputs, functionality)
### Prepare and return outputs of function
P = apollo_prepareProb(P, apollo_inputs, functionality)
return(P)
}
# ################################################################# #
#### MODEL ESTIMATION ##
# ################################################################# #
# estimate model with bfgs algorithm
mxl_wtp_case_d_rentX = apollo_estimate(apollo_beta, apollo_fixed,
apollo_probabilities, apollo_inputs,
estimate_settings=list(maxIterations=400,
estimationRoutine="bfgs",
hessianRoutine="analytic"))
# ################################################################# #
#### MODEL OUTPUTS ##
# ################################################################# #
apollo_saveOutput(mxl_wtp_case_d_rentX)