adjust plots

Scripts/mxl/Split_samples/mxl_wtp_space_vol_not_treat.R

+#### Apollo standard script #####
+
+library(apollo) # Load apollo package 
+
+
+# Test treatment effect
+
+database <- database_full %>%
+  filter(Treatment_D == "No Info 2")
+
+  #initialize model 
+  
+  apollo_initialise()
+  
+  
+  ### Set core controls
+  apollo_control = list(
+    modelName  = "MXL_wtp Vol_Not_Treat",
+    modelDescr = "MXL wtp space Vol_Not_Treat",
+    indivID    ="id",
+    mixing     = TRUE,
+    HB= FALSE,
+    nCores     = n_cores, 
+    outputDirectory = "Estimation_results/mxl/Split_samples"
+  )
+  
+  ##### 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,
+                sig_natural = 15,
+                sig_walking = 2,
+                sig_rent = 2,
+                sig_ASC_sq = 0)
+  
+  ### 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*(b_mu_natural * Naturalness_1 + b_mu_walking * WalkingDistance_1 -
+                              Rent_1)
+    
+    V[['alt2']] = -b_mu_rent*(b_mu_natural * Naturalness_2 + b_mu_walking * WalkingDistance_2 -
+                              Rent_2)
+    
+    V[['alt3']] = -b_mu_rent*(b_ASC_sq + b_mu_natural * Naturalness_3 + b_mu_walking * 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_Vol_Not_Treat = apollo_estimate(apollo_beta, apollo_fixed,
+                        apollo_probabilities, apollo_inputs, 
+                        estimate_settings=list(maxIterations=400,
+                                               estimationRoutine="bgw",
+                                               hessianRoutine="analytic"))
+  
+  
+  
+  # ################################################################# #
+  #### MODEL OUTPUTS                                               ##
+  # ################################################################# #
+  apollo_saveOutput(mxl_wtp_Vol_Not_Treat)
+
+

Scripts/mxl/Split_samples/mxl_wtp_space_vol_treat.R

@@ -123,7 +123,7 @@ database <- database_full %>%
   mxl_wtp_Vol_Treat = apollo_estimate(apollo_beta, apollo_fixed,
                         apollo_probabilities, apollo_inputs, 
                         estimate_settings=list(maxIterations=400,
-                                               estimationRoutine="bfgs",
+                                               estimationRoutine="bgw",
                                                hessianRoutine="analytic"))
   
   

Scripts/visualize_distr_spli.R

@@ -6,6 +6,8 @@ MXL_wtp_Opt_Pred_model <- apollo_loadModel("Estimation_results/mxl/prediction/Sp
 MXL_wtp_Treated_Not_Pred_model <- apollo_loadModel("Estimation_results/mxl/prediction/Split_samples/MXL_wtp_Treated_Not_Pred")
 MXL_wtp_Treated_Pred_model <- apollo_loadModel("Estimation_results/mxl/prediction/Split_samples/MXL_wtp_Treated_Pred")
 
+MXL_wtp_opt_treated <- apollo_loadModel("Estimation_results/mxl/Split_samples/MXL_wtp Vol_Treat")
+MXL_wtp_opt_not_treated <- apollo_loadModel("Estimation_results/mxl/Split_samples/MXL_wtp Vol_Not_Treat")
 
 mxl_tr <- apollo_loadModel("Estimation_results/mxl/Split_samples/MXL_wtp Treated A")
 mxl_vol_tr <- apollo_loadModel("Estimation_results/mxl/Split_samples/MXL_wtp Vol_Treated A")
@@ -18,6 +20,8 @@ models <- list(
   "Control Pred" = MXL_wtp_Control_Pred_model,
   "Opt Not Pred" = MXL_wtp_Opt_Not_Pred_model,
   "Opt Pred" = MXL_wtp_Opt_Pred_model,
+  "Opt No Info" = MXL_wtp_opt_not_treated,
+  "Opt Info" = MXL_wtp_opt_treated,
   "Treated Pred" = MXL_wtp_Treated_Pred_model,
   "Control Split Sample" = mxl_not_tr,
   "Opt Split Sample" = mxl_vol_tr,
@@ -45,6 +49,8 @@ group_mapping <- c(
   "Treated Pred" = "Treated",
   "Opt Not Pred" = "Optional",
   "Opt Pred" = "Optional",
+  "Opt No Info" = "Optional",
+  "Opt Info" = "Optional",
   "Control Split Sample" = "Control",
   "Opt Split Sample" = "Optional",
   "Treated Split Sample" = "Treated"
@@ -60,7 +66,9 @@ prediction_mapping <- c(
   "Treated Split Sample" = "Split Sample",
   "Opt Not Pred" = "Not Predicted",
   "Opt Pred" = "Predicted",
-  "Opt Split Sample" = "Split Sample"
+  "Opt Split Sample" = "Split Sample",
+  "Opt Info" = "Predicted",
+  "Opt No Info" = "Not Predicted"
 )
 
 # Define the x-range for plotting (1000 values)
@@ -128,7 +136,7 @@ calculate_z_value <- function(model_1, model_2) {
 }
 
 # Example of using the function
-z_value <- calculate_z_value(MXL_wtp_Treated_Pred_model, MXL_wtp_Control_Pred_model)
+z_value <- calculate_z_value(mxl_tr, mxl_vol_tr)
 
 # Print the z-value
 z_value
@@ -198,6 +206,14 @@ for (i in seq_along(models)) {
 ci_data$Group <- factor(ci_data$Group, levels = c("Treated", "Optional", "Control"))
 ci_data$Model <- factor(ci_data$Model, levels = unique(ci_data$Model))
 
+model_order <- c(
+  "Treated Not Pred", "Treated Pred", "Treated Split Sample",
+  "Opt Not Pred", "Opt Pred", "Opt No Info", "Opt Info", "Opt Split Sample",
+  "Control Not Pred", "Control Pred", "Control Split Sample"
+)
+ci_data$Model <- factor(ci_data$Model, levels = model_order)
+
+
 ggplot(ci_data, aes(x = Model, y = Mean, color = Group)) +
   geom_point(size = 3, position = position_dodge(width = 0.5)) +  # Mean points
   geom_errorbar(aes(ymin = Lower_CI, ymax = Upper_CI), width = 0.2, position = position_dodge(width = 0.5)) +  # CI bars