first commit

.gitignore

+.Rproj.user
+.Rhistory
+.RData
+.Ruserdata
+
+/.quarto/

choicesets.qmd

+---
+title: "Choice Sets All respondents"
+format:
+  html:
+    self-contained: true
+---
+
+
+
+```{r, echo=FALSE, include=FALSE}
+source(file = "code/feedadddata.R")
+```
+
+# All choices
+
+Summary of choice of all respondents
+
+```{r, results='asis', echo=FALSE}
+
+
+  choices %>%
+  imap(~ {
+    output <- kable(.x, caption = paste0("Choice Set Number: ", .y) ,
+                    col.names = c("Attribute", paste0("Alt ", 2:ncol(.x)-1) ) ,format = "html")%>%
+      kable_styling(bootstrap_options = "striped", full_width = FALSE) 
+    
+    cat( rep("-", 30), "\n", output , " \n Choice Sets ends") 
+  }
+  
+  )
+
+
+```
+
+
+
+
+

choicesets_nonattendence.qmd

+---
+title: "Choice Sets Attribute Non-Attendence"
+format:
+  html:
+    self-contained: true
+---
+
+
+
+```{r, echo=FALSE, include=FALSE}
+source(file = "code/feedadddata.R")
+```
+
+# Nonattendence
+
+Summary of choice of all respondents
+
+```{r, results='asis', echo=FALSE, warning=FALSE, message=FALSE}
+
+
+  allheuristics[[2]] %>%
+  imap(~ {
+    output <- kable(.x, caption = paste0("Choice Set Number: ", .y) ,
+                    col.names = c("Attribute", paste0("Alt ", 2:ncol(.x)-1) ) ,format = "html")%>%
+      kable_styling(bootstrap_options = "striped", full_width = FALSE) 
+    
+    cat( rep("-", 30), "\n", output , " \n Choice Sets ends") 
+  }
+  
+  )
+
+
+```
+
+
+
+
+

choicesets_priceheuristic.qmd

+---
+title: "Choice Sets Price Heuristics"
+format:
+  html:
+    toc: true
+---
+
+
+
+```{r, echo=FALSE, include=FALSE}
+source(file = "code/feedadddata.R")
+```
+
+# Price Heuristic
+
+Summary of choice of all respondents
+
+```{r, results='asis', echo=FALSE}
+
+
+  allheuristics[[3]] %>%
+  imap(~ {
+    output <- kable(.x, caption = paste0("Choice Set Number: ", .y) ,
+                    col.names = c("Attribute", paste0("Alt ", 2:ncol(.x)-1) ) ,format = "html")%>%
+      kable_styling(bootstrap_options = "striped", full_width = FALSE) 
+    
+    cat( rep("-", 30), "\n", output , " \n Choice Sets ends") 
+  }
+  
+  )
+
+
+```
+
+
+
+
+

choicesets_rum.qmd

+---
+title: "Choice Sets RUM"
+format:
+  html:
+    toc: true
+---
+
+
+
+```{r, echo=FALSE, include=FALSE}
+source(file = "code/feedadddata.R")
+```
+
+# RUM
+
+Summary of choice of all respondents
+
+```{r, results='asis', echo=FALSE}
+
+
+  allheuristics[[1]] %>%
+  imap(~ {
+    output <- kable(.x, caption = paste0("Choice Set Number: ", .y) ,
+                    col.names = c("Attribute", paste0("Alt ", 2:ncol(.x)-1) ) ,format = "html")%>%
+      kable_styling(bootstrap_options = "striped", full_width = FALSE) 
+    
+    cat( rep("-", 30), "\n", output , " \n Choice Sets ends") 
+  }
+  
+  )
+
+
+```
+
+
+
+
+

code/3lc_nocov.R

+
+rm(list = ls(, pattern = "^apollo_"))
+
+database <- feedadd_data
+
+apollo_initialise()
+
+### Set core controls
+apollo_control = list(
+  modelName  ="lc3cl",
+  modelDescr ="3 Class LCL model for Feed Additives DCE",
+  indivID    ="ID",
+  nCores     = 11 , 
+  outputDirectory = "modeloutput/lclogit/"
+)
+
+# ################################################################# #
+#### DEFINE MODEL PARAMETERS                                     ####
+# ################################################################# #
+
+### Vector of parameters, including any that are kept fixed in estimation
+
+lcnames <- function(apollobeta, classnames) {
+  
+  x<-NULL
+  
+  for (i in classnames) {
+    g <- gsub("$", paste0("_",i),names(apollo_beta))
+    x<-c(x,g)
+  }
+  
+  apollo_beta <- rep(apollo_beta, length(classnames))
+  
+  names(apollo_beta) <- x 
+  
+  return(apollo_beta)
+  
+}
+
+
+apollo_beta <-c(basc = 0.2,
+                bcow = 0.3,
+                badv = 0.3,
+                bvet = 0.3,
+                bfar = 0.3,
+                bmet = 0.3,
+                bbon = 0.3,
+                delta = 0
+)
+
+
+
+
+
+
+
+
+apollo_beta<-lcnames(apollo_beta, c("a","b","c"))
+
+
+
+### Vector with names (in quotes) of parameters to be kept fixed at their starting value in apollo_beta, use apollo_beta_fixed = c() if none
+apollo_fixed = c("delta_a")
+
+
+# ################################################################# #
+#### DEFINE LATENT CLASS COMPONENTS                              ####
+# ################################################################# #
+
+apollo_lcPars=function(apollo_beta, apollo_inputs){
+  
+  
+  
+  
+  lcpars = list()
+  
+  
+  
+  lcpars[["basc"]] = list(basc_a , basc_b, basc_c)
+  lcpars[["bcow"]] = list(bcow_a, bcow_b ,bcow_c )
+  lcpars[["badv"]] = list(badv_a, badv_b ,badv_c)
+  lcpars[["bvet"]]  = list(bvet_a, bvet_b , bvet_c)
+  lcpars[["bfar"]]  = list(bfar_a, bfar_b , bfar_c)
+  lcpars[["bmet"]] = list(bmet_a, bmet_b , bmet_c)
+  lcpars[["bbon"]] = list(bbon_a, bbon_b, bbon_c)
+
+  
+  V=list()
+  V[["class_a"]] = delta_a
+  V[["class_b"]] = delta_b
+  V[["class_c"]] = delta_c
+  
+  mnl_settings = list(
+    alternatives = c(class_a=1, class_b=2, class_c=3), 
+    avail        = 1, 
+    choiceVar    = NA, 
+    V            = V
+  )
+  
+  lcpars[["pi_values"]] = apollo_mnl(mnl_settings, functionality="raw")
+  
+  lcpars[["pi_values"]] = apollo_firstRow(lcpars[["pi_values"]], apollo_inputs)
+  
+  return(lcpars)
+}
+
+
+# ################################################################# #
+#### GROUP AND VALIDATE INPUTS                                   ####
+# ################################################################# #
+
+apollo_inputs = apollo_validateInputs()
+
+# ################################################################# #
+#### DEFINE MODEL AND LIKELIHOOD FUNCTION                        ####
+# ################################################################# #
+
+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()
+  
+  
+  
+  ### Define settings for MNL model component that are generic across classes
+  mnl_settings = list(
+    alternatives = c(alt1=1, alt2=2, alt3=3),
+    avail        = 1,
+    choiceVar    = CHOICE
+  )
+  
+  
+  ### Loop over classes
+  
+  
+  for (s in 1:3) {
+    
+    ### Compute class-specific utilities
+     V=list()
+    
+    
+    V[['alt1']] = basc[[s]] + bbon[[s]]*alt1_bon +  badv[[s]]*alt1_adv + bcow[[s]]*alt1_cow + bvet[[s]]*alt1_vet + bfar[[s]]*alt1_vet +  bmet[[s]]*alt1_met 
+    V[['alt2']] = basc[[s]] + bbon[[s]]*alt2_bon +  badv[[s]]*alt2_adv + bcow[[s]]*alt2_cow + bvet[[s]]*alt2_vet + bfar[[s]]*alt2_vet  + bmet[[s]]*alt2_met    
+    V[['alt3']] = 0 
+    
+    
+    mnl_settings$V = V
+    mnl_settings$componentName = paste0("Class_",s)
+    
+    ### Compute within-class choice probabilities using MNL model
+    P[[paste0("Class_",s)]] = apollo_mnl(mnl_settings, functionality)
+    
+    ### Take product across observation for same individual
+    P[[paste0("Class_",s)]] = apollo_panelProd(P[[paste0("Class_",s)]], apollo_inputs ,functionality)
+    
+    
+  }
+  
+  ### Compute latent class model probabilities
+  lc_settings   = list(inClassProb = P, classProb=pi_values)
+  P[["model"]] = apollo_lc(lc_settings, apollo_inputs, functionality)
+  
+  ### Prepare and return outputs of function
+  P = apollo_prepareProb(P, apollo_inputs, functionality)
+  return(P)
+} 
+
+
+lc3cl = apollo_estimate(apollo_beta, apollo_fixed,
+                        apollo_probabilities, apollo_inputs, 
+                        estimate_settings=list(hessianRoutine="maxLik"))
+
+
+lc3cl[["unconditionals"]] <- apollo_lcUnconditionals(lc3cl,apollo_probabilities,apollo_inputs)
+
+apollo_saveOutput(lc3cl,  saveOutput_settings = list(saveEst=FALSE, savbmetv=FALSE, savbmetrr=FALSE))
+
+
+
+
+
+

code/feedadddata.R

+rm(list=ls())
+#remotes::install_gitlab(repo = "dj44vuri/choicetools" , host = "https://git.idiv.de")
+library(dplyr)
+library(choiceTools)
+library(kableExtra)
+library(purrr)
+library("texreg")
+
+
+
+feedadd_data <- readRDS("data/feedadd_data.RDS")
+
+
+table(feedadd_data$group)
+
+table(feedadd_data$group)/(length(feedadd_data$group))
+
+
+
+choices <- choiceTools::createSets(
+  feedadd_data,choice = "CHOICE", uniquerow = "Choice_situation",
+  attributes = starts_with("alt"), prefix = "alt"
+    )
+
+allheuristics <-map(seq_along(unique(feedadd_data$group)), 
+  ~  createSets(
+  feedadd_data %>% filter(group==.x) ,choice = "CHOICE", uniquerow = "Choice_situation",
+  attributes = starts_with("alt"), prefix = "alt"
+) )
+
+
+
+
+
+#### Small data set
+
+smalldata <- feedadd_data %>% 
+  distinct(Choice_situation, group, .keep_all = TRUE) %>% 
+  select(!matches("[1-3]$") ,-Block, -group ) 
+
+mnlsplits <- list()
+splitnames <- c("ALL", "RUM" , "NonAttendence" , "Price")
\ No newline at end of file

code/readmodels.R

+
+deleteold <-list.files("modeloutput/clogit/", pattern = "OLD", full.names=TRUE)
+
+file.remove(deleteold)
+
+mnlsplits<-readRDS("modeloutput/modelstex.RDS")
+
+
+lc3cl_model <- readRDS("~/share/groups/bioecon/Julian/seminarheuristics/modeloutput/lclogit/lc3cl_model.rds")
\ No newline at end of file

code/simplemnl_body.R

+apollo_beta <-c(basc = 0.2,
+                bcow = 0.3,
+                badv = 0.3,
+                bvet = 0.3,
+                bfar = 0.3,
+                bmet = 0.3,
+                bbon = 0.3
+                )
+
+
+### keine Parameter fix halten
+apollo_fixed <- c()
+
+### validieren
+#assign("apollo_beta",apollo_beta, envir = .GlobalEnv)
+#assign("apollo_fixed",apollo_fixed, envir = .GlobalEnv)
+#assign("apollo_control",apollo_control, envir = .GlobalEnv)
+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)
+  V = list()
+  V[['alt1']] = bcow*alt1_cow + badv * alt1_adv + bvet * alt1_vet + bfar * alt1_far + bmet*alt1_met + bbon * alt1_bon 
+  V[['alt2']] = bcow*alt2_cow + badv * alt2_adv + bvet * alt2_vet + bfar * alt2_far + bmet*alt2_met + bbon * alt2_bon    
+  V[['alt3']] = basc
+  
+
+  
+  ### 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)
+}
+
+
+
+
+

code/simplemnl_split.R

+library(apollo)
+database <- feedadd_data
+mnlsplits <- list()
+splitnames <- c("ALL", "RUM" , "NonAttendence" , "Price")
+
+
+for (splitno in seq_along(1:length(splitnames))) {
+
+  print(splitnames[splitno])
+    
+
+
+if (splitno!=1) {
+  database<-feedadd_data[which(feedadd_data$group==splitno-1),] 
+}  
+
+apollo_initialise()
+
+
+apollo_control <- list(
+  modelName  = paste0("MNL_",splitnames[splitno]),
+  modelDescr =paste0("MNL model Split ",splitnames[splitno]),
+  indivID    ="ID", 
+  outputDirectory = "modeloutput/clogit/"
+)
+
+source("code/simplemnl_body.R")
+
+
+
+m <- apollo_estimate(apollo_beta, apollo_fixed,
+                          apollo_probabilities, apollo_inputs, 
+                          estimate_settings=list(hessianRoutine="maxLik"))
+
+apollo_saveOutput(m,  saveOutput_settings = list(saveEst=FALSE, saveCov=FALSE, saveCorr=FALSE))
+
+mnlsplits[[splitnames[splitno]]]<-quicktexregapollo(m)
+
+}
+
+ 
+
+saveRDS(mnlsplits,file = "modeloutput/modelstex.RDS")
+

modeloutput/clogit/MNL_ALL_iterations.csv

+"basc","bcow","badv","bvet","bfar","bmet","bbon","logLike"
+0.2,0.3,0.3,0.3,0.3,0.3,0.3,-20713.3421930812
+0.2,0.3,0.3,0.3,0.3,0.3,0.3,-20713.3421930812
+0.197774283346829,0.29993519713499,0.304012938177298,0.300431413218036,0.296521037378937,0.274274026456621,0.366368731354262,-20485.7925398724
+0.209055461544678,0.289586361312179,0.307510722969549,0.302411348243303,0.284572684268648,0.147998222913648,0.361166935851261,-20436.5003117349
+0.19784599933696,0.290293804258131,0.329575131751255,0.331241661657545,0.255306889485088,0.14683906545105,0.358668502201278,-20435.1755400108
+0.190327002382416,0.286327253232415,0.330080671333177,0.33484784581644,0.258843338079719,0.146090883285112,0.358310678882282,-20434.9766063357
+0.18434307624573,0.283075801427147,0.3297253167463,0.332075277225079,0.256768923942102,0.145931245358543,0.357998919975628,-20434.8567878365
+0.176521942211385,0.28543982790769,0.326685870599895,0.330227740910902,0.254670673736218,0.145121158711483,0.357530278314043,-20434.7365586164
+0.174056302945655,0.288268650136706,0.331242079102149,0.327430513911481,0.256311815034111,0.144717946090073,0.357124734600501,-20434.665267208
+0.17389577488282,0.288241694821458,0.325715088561486,0.322685611206614,0.261772167291991,0.163184669461013,0.364683782166024,-20411.6283139804
+0.167883979396811,0.287201261090668,0.317295549838884,0.316571514401408,0.266782165061282,0.183435346801448,0.373507729061407,-20402.0087562361
+0.1673920244584,0.287096532541116,0.317241422066306,0.316620428882487,0.267325831621783,0.18407466167109,0.373974808023803,-20401.9908586195
+0.167339335675083,0.287093008465451,0.317262142229566,0.316649173321282,0.267350781049214,0.184073677307018,0.37399349029677,-20401.990830438
+0.167339310681464,0.287093020337431,0.317262161361498,0.316649187678528,0.267350793284478,0.1840737009989,0.373993664068592,-20401.9908304335
+0.167339258710521,0.287093087478575,0.317262501679274,0.316649509090992,0.267350826875124,0.184073386918355,0.373994565021228,-20401.9908304287

modeloutput/clogit/MNL_ALL_output.txt

+Model run by dj44vuri using Apollo 0.2.8 on R 4.2.2 for Linux.
+www.ApolloChoiceModelling.com
+
+Model name                                  : MNL_ALL
+Model description                           : MNL model Split ALL
+Model run at                                : 2023-11-27 23:36:39
+Estimation method                           : bfgs
+Model diagnosis                             : successful convergence 
+Number of individuals                       : 3600
+Number of rows in database                  : 28800
+Number of modelled outcomes                 : 28800
+
+Number of cores used                        :  1 
+Model without mixing
+
+LL(start)                                   : -20713.34
+LL at equal shares, LL(0)                   : -31640.03
+LL at observed shares, LL(C)                : -25327.01
+LL(final)                                   : -20401.99
+Rho-squared vs equal shares                  :  0.3552 
+Adj.Rho-squared vs equal shares              :  0.355 
+Rho-squared vs observed shares               :  0.1945 
+Adj.Rho-squared vs observed shares           :  0.1942 
+AIC                                         :  40817.98 
+BIC                                         :  40875.86 
+
+Estimated parameters                        :  7
+Time taken (hh:mm:ss)                       :  00:00:5.67 
+     pre-estimation                         :  00:00:1.5 
+     estimation                             :  00:00:2.03 
+     post-estimation                        :  00:00:2.14 
+Iterations                                  :  15  
+Min abs eigenvalue of Hessian               :  541.3482 
+
+Unconstrained optimisation.
+
+These outputs have had the scaling
+  used in estimation applied to
+  them.
+Estimates:
+        Estimate        s.e.
+basc      0.1673    0.040025
+bcow      0.2871    0.018514
+badv      0.3173    0.016217
+bvet      0.3166    0.018218
+bfar      0.2674    0.016493
+bmet      0.1841    0.007733
+bbon      0.3740    0.004567
+       t.rat.(0)    Rob.s.e.
+basc       4.181    0.039099
+bcow      15.507    0.017793
+badv      19.563    0.015581
+bvet      17.381    0.017360
+bfar      16.209    0.016172
+bmet      23.803    0.008632
+bbon      81.884    0.005333
+     Rob.t.rat.(0)
+basc         4.280
+bcow        16.135
+badv        20.362
+bvet        18.240
+bfar        16.532
+bmet        21.325
+bbon        70.128
+
+
+Overview of choices for MNL model
+  component :
+                                     alt1
+Times available                  28800.00
+Times chosen                     13067.00
+Percentage chosen overall           45.37
+Percentage chosen when available    45.37
+                                     alt2
+Times available                  28800.00
+Times chosen                     13985.00
+Percentage chosen overall           48.56
+Percentage chosen when available    48.56
+                                     alt3
+Times available                  28800.00
+Times chosen                      1748.00
+Percentage chosen overall            6.07
+Percentage chosen when available     6.07
+
+
+Classical covariance matrix:
+            basc        bcow
+basc    0.001602  1.7961e-04
+bcow  1.7961e-04  3.4278e-04
+badv  2.6768e-04  -2.369e-05
+bvet  3.2945e-04  -3.743e-06
+bfar  3.1445e-04   7.377e-06
+bmet  1.2799e-04  -5.287e-06
+bbon   9.600e-05   4.210e-06
+            badv        bvet
+basc  2.6768e-04  3.2945e-04
+bcow  -2.369e-05  -3.743e-06
+badv  2.6300e-04  1.1377e-04
+bvet  1.1377e-04  3.3191e-04
+bfar   8.777e-05  1.0934e-04
+bmet   2.178e-05   2.172e-05
+bbon   7.781e-06   1.751e-05
+            bfar        bmet
+basc  3.1445e-04  1.2799e-04
+bcow   7.377e-06  -5.287e-06
+badv   8.777e-05   2.178e-05
+bvet  1.0934e-04   2.172e-05
+bfar  2.7204e-04   2.412e-05
+bmet   2.412e-05   5.980e-05
+bbon   2.032e-05   3.745e-06
+            bbon
+basc   9.600e-05
+bcow   4.210e-06
+badv   7.781e-06
+bvet   1.751e-05
+bfar   2.032e-05
+bmet   3.745e-06
+bbon   2.086e-05
+
+Robust covariance matrix:
+            basc        bcow
+basc    0.001529  1.9358e-04
+bcow  1.9358e-04  3.1659e-04
+badv  2.4715e-04  -1.421e-05
+bvet  3.1069e-04   3.207e-06
+bfar  3.3138e-04   2.351e-05
+bmet  1.4444e-04   7.703e-07
+bbon   9.105e-05   4.678e-06
+            badv        bvet
+basc  2.4715e-04  3.1069e-04
+bcow  -1.421e-05   3.207e-06
+badv  2.4276e-04  1.1289e-04
+bvet  1.1289e-04  3.0138e-04
+bfar   8.859e-05  1.1381e-04
+bmet   1.976e-05   2.370e-05
+bbon   8.444e-06   1.886e-05
+            bfar        bmet
+basc  3.3138e-04  1.4444e-04
+bcow   2.351e-05   7.703e-07
+badv   8.859e-05   1.976e-05
+bvet  1.1381e-04   2.370e-05
+bfar  2.6152e-04   3.661e-05
+bmet   3.661e-05   7.451e-05
+bbon   2.239e-05   3.970e-07
+            bbon
+basc   9.105e-05
+bcow   4.678e-06
+badv   8.444e-06
+bvet   1.886e-05
+bfar   2.239e-05
+bmet   3.970e-07
+bbon   2.844e-05
+
+Classical correlation matrix:
+            basc        bcow
+basc      1.0000     0.24238
+bcow      0.2424     1.00000
+badv      0.4124    -0.07890
+bvet      0.4518    -0.01110
+bfar      0.4763     0.02416
+bmet      0.4135    -0.03693
+bbon      0.5251     0.04978
+            badv        bvet
+basc     0.41239     0.45180
+bcow    -0.07890    -0.01110
+badv     1.00000     0.38507
+bvet     0.38507     1.00000
+bfar     0.32815     0.36387
+bmet     0.17366     0.15419
+bbon     0.10505     0.21041
+            bfar        bmet
+basc     0.47634     0.41352
+bcow     0.02416    -0.03693
+badv     0.32815     0.17366
+bvet     0.36387     0.15419
+bfar     1.00000     0.18907
+bmet     0.18907     1.00000
+bbon     0.26975     0.10602
+            bbon
+basc     0.52515
+bcow     0.04978
+badv     0.10505
+bvet     0.21041
+bfar     0.26975
+bmet     0.10602
+bbon     1.00000
+
+Robust correlation matrix:
+            basc        bcow
+basc      1.0000    0.278254
+bcow      0.2783    1.000000
+badv      0.4057   -0.051275
+bvet      0.4577    0.010381
+bfar      0.5241    0.081702
+bmet      0.4280    0.005016
+bbon      0.4367    0.049295
+            badv        bvet
+basc     0.40570     0.45772
+bcow    -0.05127     0.01038
+badv     1.00000     0.41735
+bvet     0.41735     1.00000
+bfar     0.35158     0.40537
+bmet     0.14692     0.15814
+bbon     0.10163     0.20375
+            bfar        bmet
+basc     0.52409    0.427963
+bcow     0.08170    0.005016
+badv     0.35158    0.146916
+bvet     0.40537    0.158142
+bfar     1.00000    0.262244
+bmet     0.26224    1.000000
+bbon     0.25964    0.008623
+            bbon
+basc    0.436674
+bcow    0.049295
+badv    0.101626
+bvet    0.203748
+bfar    0.259644
+bmet    0.008623
+bbon    1.000000
+
+ 20 worst outliers in terms of lowest average per choice prediction:
+   ID Avg prob per choice
+ 2245           0.1291638
+ 2459           0.1361100
+ 2674           0.1445860
+ 2660           0.1559322
+ 2755           0.1608655
+  513           0.1660440
+ 2632           0.1738870
+ 2643           0.1757546
+ 2456           0.1764956
+  504           0.1796079
+  626           0.1837007
+ 2770           0.1888183
+ 1455           0.1901808
+ 2614           0.1940656
+  127           0.1954228
+  676           0.1956396
+ 2665           0.1968922
+ 1174           0.1970817
+ 1668           0.1981441
+ 2580           0.1988892
+
+Changes in parameter estimates from starting values:
+         Initial    Estimate
+basc      0.2000      0.1673
+bcow      0.3000      0.2871
+badv      0.3000      0.3173
+bvet      0.3000      0.3166
+bfar      0.3000      0.2674
+bmet      0.3000      0.1841
+bbon      0.3000      0.3740
+      Difference
+basc    -0.03266
+bcow    -0.01291
+badv     0.01726
+bvet     0.01665
+bfar    -0.03265
+bmet    -0.11593
+bbon     0.07399
+
+Settings and functions used in model definition:
+
+apollo_control
+--------------
+                       Value                
+modelName              "MNL_ALL"            
+modelDescr             "MNL model Split ALL"
+indivID                "ID"                 
+outputDirectory        "modeloutput/clogit/"
+debug                  "FALSE"              
+nCores                 "1"                  
+workInLogs             "FALSE"              
+seed                   "13"                 
+mixing                 "FALSE"              
+HB                     "FALSE"              
+noValidation           "FALSE"              
+noDiagnostics          "FALSE"              
+calculateLLC           "TRUE"               
+panelData              "TRUE"               
+analyticGrad           "TRUE"               
+analyticGrad_manualSet "FALSE"              
+
+Hessian routines attempted
+--------------
+numerical second derivative of LL (using maxLik)
+
+Scaling in estimation
+--------------
+         Value
+basc 0.1673393
+bcow 0.2870930
+badv 0.3172622
+bvet 0.3166492
+bfar 0.2673508
+bmet 0.1840737
+bbon 0.3739937
+
+Scaling used in computing Hessian
+--------------
+         Value
+basc 0.1673393
+bcow 0.2870931
+badv 0.3172625
+bvet 0.3166495
+bfar 0.2673508
+bmet 0.1840734
+bbon 0.3739946
+
+
+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)
+  V = list()
+  V[['alt1']] = bcow*alt1_cow + badv * alt1_adv + bvet * alt1_vet + bfar * alt1_far + bmet*alt1_met + bbon * alt1_bon 
+  V[['alt2']] = bcow*alt2_cow + badv * alt2_adv + bvet * alt2_vet + bfar * alt2_far + bmet*alt2_met + bbon * alt2_bon    
+  V[['alt3']] = basc
+  
+
+  
+  ### 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)
+}
+<bytecode: 0x5574d4c15958>

modeloutput/clogit/MNL_NonAttendence_iterations.csv

+"basc","bcow","badv","bvet","bfar","bmet","bbon","logLike"
+0.2,0.3,0.3,0.3,0.3,0.3,0.3,-5044.84843861625
+0.2,0.3,0.3,0.3,0.3,0.3,0.3,-5044.84843861625
+0.212263883346829,0.29179439713499,0.299968138177298,0.296668213218036,0.297749837378937,0.206882026456621,0.246355931354261,-4915.8521679726
+0.21505322247987,0.288007052371971,0.307066931294511,0.296398353764042,0.296203693902158,0.128184562018626,0.284825156983034,-4855.40348760455
+0.164810465913807,0.302835493148867,0.404816540314275,0.356489206697806,0.318249158638829,0.116318588513749,0.274286784742645,-4853.52074844387
+0.151927734895124,0.279975000579782,0.400172090518385,0.371946082312472,0.32381086785241,0.115431130753391,0.27414457358128,-4853.14694507272
+0.103543251112594,0.240604976106548,0.413191612897785,0.333584953609224,0.318095463077274,0.113847993951548,0.271528815115277,-4853.13129005344
+-0.0257209506573074,0.252349339629921,0.374297025725588,0.305970791318832,0.284610846300238,0.0974347955791483,0.263859650286435,-4851.58567194988
+0.0155565738945079,0.237742546447694,0.386607518489969,0.327029281664056,0.205849152706199,0.0800326994725245,0.266140143478844,-4843.09430085209
+0.0818548031583402,0.247102649773792,0.379395400602803,0.321914365858315,0.263119248631678,0.0420350264918261,0.271138755397524,-4823.71571932239
+0.164348115536451,0.264697962443476,0.367801882288876,0.330938686649523,0.308001619013226,0.0035633459370288,0.281561672447437,-4817.40456579944
+0.1646237609054,0.265554058018546,0.368510399528634,0.332676792467705,0.309976604653833,0.00304292134497875,0.282269917085986,-4817.39657694711
+0.165080019525179,0.265622433753292,0.368460867232662,0.332563361113081,0.309883881109384,0.00308300487879505,0.282260154619513,-4817.39649828911
+0.165041463437502,0.265622032920252,0.368455655796549,0.332553897406415,0.309879698263003,0.00307878099338118,0.282256788431256,-4817.39649815984
+0.165041464214499,0.265622031930734,0.368455676310814,0.33255390929899,0.309879693524504,0.00307878096775442,0.282256858465724,-4817.39649815968