Model run by trashe using Apollo 0.4.0 on R 4.5.3 for Linux. Please acknowledge the use of Apollo by citing Hess & Palma (2019) DOI 10.1016/j.jocm.2019.100170 www.ApolloChoiceModelling.com Model name : MNL_own_model Model description : Binary logit on Swiss route choice data, written as a user-defined model with apollo_ownModel Model run at : 2026-11-02 21:58:40 Estimation method : bgw Estimation diagnosis : Relative function convergence (favorable) Optimisation diagnosis : Maximum found hessian properties : Negative definite maximum eigenvalue : -527.597401 reciprocal of condition number : 0.00157623 Number of individuals : 388 Number of rows in database : 3492 Number of modelled outcomes : 3492 Number of cores used : 1 Model without mixing LL(start) : -2420.47 LL at equal shares, LL(0) : -2420.47 LL at observed shares, LL(C) : -2420.39 LL(final) : -1665.62 Rho-squared vs equal shares : 0.3119 Adj.Rho-squared vs equal shares : 0.3098 Rho-squared vs observed shares : 0.3118 Adj.Rho-squared vs observed shares : Not applicable AIC : 3341.24 BIC : 3372.03 Estimated parameters : 5 Time taken (hh:mm:ss) : 00:00:0.82 pre-estimation : 00:00:0.68 estimation : 00:00:0.09 post-estimation : 00:00:0.05 Iterations : 10 Unconstrained optimisation. Estimates: Estimate s.e. t.rat.(0) Rob.s.e. Rob.t.rat.(0) asc_1 -0.01587 0.042870 -0.3703 0.045658 -0.3477 b_tt -0.05975 0.004257 -14.0359 0.006744 -8.8606 b_tc -0.13173 0.013505 -9.7545 0.023641 -5.5721 b_hw -0.03745 0.001848 -20.2681 0.002317 -16.1593 b_ch -1.15212 0.043420 -26.5343 0.061367 -18.7743 Value of travel time: 27.22 CHF/hour. Route 1 chosen in 49.7% of cases. Classical covariance matrix: asc_1 b_tt b_tc b_hw b_ch asc_1 0.001838 -1.833e-06 -8.893e-06 -2.855e-07 1.222e-05 b_tt -1.833e-06 1.812e-05 4.574e-05 1.643e-06 5.001e-05 b_tc -8.893e-06 4.574e-05 1.8238e-04 4.109e-06 1.0047e-04 b_hw -2.855e-07 1.643e-06 4.109e-06 3.413e-06 2.598e-05 b_ch 1.222e-05 5.001e-05 1.0047e-04 2.598e-05 0.001885 Robust covariance matrix: asc_1 b_tt b_tc b_hw b_ch asc_1 0.002085 -3.172e-05 -1.0166e-04 -6.807e-06 -1.5887e-04 b_tt -3.172e-05 4.548e-05 1.1770e-04 2.934e-06 1.3779e-04 b_tc -1.0166e-04 1.1770e-04 5.5891e-04 6.570e-06 3.1532e-04 b_hw -6.807e-06 2.934e-06 6.570e-06 5.370e-06 4.953e-05 b_ch -1.5887e-04 1.3779e-04 3.1532e-04 4.953e-05 0.003766 Classical correlation matrix: asc_1 b_tt b_tc b_hw b_ch asc_1 1.000000 -0.01004 -0.01536 -0.003605 0.006566 b_tt -0.010043 1.00000 0.79562 0.208850 0.270564 b_tc -0.015361 0.79562 1.00000 0.164664 0.171332 b_hw -0.003605 0.20885 0.16466 1.000000 0.323873 b_ch 0.006566 0.27056 0.17133 0.323873 1.000000 Robust correlation matrix: asc_1 b_tt b_tc b_hw b_ch asc_1 1.00000 -0.1030 -0.09418 -0.06434 -0.05670 b_tt -0.10301 1.0000 0.73828 0.18774 0.33295 b_tc -0.09418 0.7383 1.00000 0.11991 0.21734 b_hw -0.06434 0.1877 0.11991 1.00000 0.34833 b_ch -0.05670 0.3330 0.21734 0.34833 1.00000 Settings and functions used in model definition: apollo_control -------------- Value modelName "MNL_own_model" modelDescr "Binary logit on Swiss route choice data, written as a user-defined model with apollo_ownModel" indivID "ID" outputDirectory "output/" debug "FALSE" nCores "1" workInLogs "FALSE" seed "13" mixing "FALSE" HB "FALSE" noValidation "FALSE" noDiagnostics "FALSE" calculateLLC "TRUE" analyticHessian "FALSE" dropLoopExpansion "FALSE" dropRedefCheck "FALSE" panelData "TRUE" analyticGrad "TRUE" analyticGrad_manualSet "FALSE" overridePanel "FALSE" preventOverridePanel "FALSE" noModification "FALSE" Hessian routines attempted -------------------------- numerical jacobian of LL analytical gradient apollo_probabilities ---------------------- function (apollo_beta, apollo_inputs, functionality = "estimate") { apollo_attach(apollo_beta, apollo_inputs) on.exit(apollo_detach(apollo_beta, apollo_inputs)) P = list() V1 = asc_1 + b_tt * tt1 + b_tc * tc1 + b_hw * hw1 + b_ch * ch1 V2 = b_tt * tt2 + b_tc * tc2 + b_hw * hw2 + b_ch * ch2 P1 = exp(V1)/(exp(V1) + exp(V2)) y1 = 1 * (choice == 1) ownModel_settings = list(likelihood = y1 * P1 + (1 - y1) * (1 - P1), zero_LL = 0.5, shares_LL = y1 * mean(y1) + (1 - y1) * (1 - mean(y1)), prediction = P1, report = list(data = paste0("Route 1 chosen in ", round(100 * mean(y1), 1), "% of cases."), param = paste0("Value of travel time: ", round(60 * b_tt/b_tc, 2), " CHF/hour."))) P[["model"]] = apollo_ownModel(ownModel_settings, functionality) P = apollo_panelProd(P, apollo_inputs, functionality) P = apollo_prepareProb(P, apollo_inputs, functionality) return(P) }