additional baseline version

input/v6.3/params/baseline_v4.yaml

@@ -0,0 +1,39 @@
+#Modes:  ['walk', 'pt', 'car', 'bike', 'ride']
+#Number of choices:  27235
+#Varying: ['km', 'hours', 'walking_km', 'switches', 'valid']
+#  Results for model trip_choice_performing_exp_income_util_money_fixed_price_perception
+#Nbr of parameters:		8
+#Sample size:			27235
+#Excluded data:			0
+#Null log likelihood:		-40155.25
+#Final log likelihood:		-31522.25
+#Likelihood ratio test (null):		17266
+#Rho square (null):			0.215
+#Rho bar square (null):			0.215
+#Akaike Information Criterion:	63060.49
+#Bayesian Information Criterion:	63126.19
+#  Value  Rob. Std err  Rob. t-test  Rob. p-value
+#  ASC_bike              -1.689491      0.032510   -51.968167  0.000000e+00
+#  ASC_car               -1.920529      0.046183   -41.584864  0.000000e+00
+#  ASC_pt                -0.866749      0.029893   -28.995219  0.000000e+00
+#  ASC_ride              -2.401215      0.033845   -70.946824  0.000000e+00
+#  BETA_PERFORMING        5.677002      0.166151    34.167734  0.000000e+00
+#  BETA_PRICE_PERCEPTION  0.216186      0.027424     7.883173  3.108624e-15
+#  EXP_INCOME             0.197429      0.036124     5.465322  4.620660e-08
+#  UTIL_MONEY             0.534645      0.028387    18.834091  0.000000e+00
+
+scoring:
+  scoringParameters:
+    - performing: 5.677002
+      marginalUtilityOfMoney: 0.534645
+      modeParams:
+        - mode: walk
+          constant: 0
+        - mode: car
+          constant: 0
+          dailyMonetaryConstant: -3.0914598
+        - mode: pt
+          constant: 0
+          dailyMonetaryConstant: -0.648558
+advancedScoring:
+  incomeExponent: 0.197429

src/main/python/choicemodels/estimate_biogeme_trip_choice.py

@@ -2,14 +2,12 @@
 # -*- coding: utf-8 -*-
 
 import argparse
-from collections import defaultdict
-
 import biogeme.biogeme as bio
 import biogeme.database as db
 import biogeme.models as models
 from biogeme.expressions import Beta, bioDraws, PanelLikelihoodTrajectory, log, MonteCarlo
 
-from prepare import read_trip_choices
+from prepare import read_trip_choices, daily_costs, km_costs
 
 ESTIMATE = 0
 FIXED = 1
@@ -22,6 +20,9 @@
     parser.add_argument("--est-performing", help="Estimate the beta for performing", action="store_true")
     parser.add_argument("--est-exp-income", help="Estimate exponent for income", action="store_true")
     parser.add_argument("--est-util-money", help="Estimate utility of money", action="store_true")
+    parser.add_argument("--est-fixed-price-perception", help="Estimate price perception of daily costs",
+                        action="store_true")
+
     parser.add_argument("--no-income", help="Don't consider the income", action="store_true")
 
     args = parser.parse_args()
@@ -34,8 +35,6 @@
     database = db.Database("data/choices", df)
     v = database.variables
 
-    km_costs = defaultdict(lambda: 0.0, car=-0.149, ride=-0.149)
-
     ASC = {}
     for mode in ds.modes:
         # Base asc
@@ -51,11 +50,14 @@
     EXP_INCOME = Beta('EXP_INCOME', 1, 0, 1.5, ESTIMATE if args.est_exp_income else FIXED)
     UTIL_MONEY = Beta('UTIL_MONEY', 1, 0, 2, ESTIMATE if args.est_util_money else FIXED)
     BETA_PERFORMING = Beta('BETA_PERFORMING', 6.88, 1, 15, ESTIMATE if args.est_performing else FIXED)
+    BETA_PRICE_PERCEPTION = Beta('BETA_PRICE_PERCEPTION', 0, 0, 1,
+                                 ESTIMATE if args.est_fixed_price_perception else FIXED)
 
     for i, mode in enumerate(ds.modes, 1):
         u = ASC[mode] - BETA_PERFORMING * v[f"{mode}_hours"] * (2 if mode == "ride" else 1)
 
         price = km_costs[mode] * v[f"{mode}_km"]
+        price += daily_costs[mode] * v["dist_weight"] * BETA_PRICE_PERCEPTION
         u += price * UTIL_MONEY * (1 if args.no_income else (ds.global_income / v["income"]) ** EXP_INCOME)
 
         if mode == "pt":
@@ -83,6 +85,8 @@
         modelName += "_exp_income"
     if args.est_util_money:
         modelName += "_util_money"
+    if args.est_fixed_price_perception:
+        modelName += "_fixed_price_perception"
 
     biogeme.modelName = modelName
     biogeme.weight = v["weight"]

src/main/python/choicemodels/prepare.py

@@ -142,4 +142,10 @@ def read_trip_choices(input_file: str) -> TripChoice:
 
     print("Varying:", varying)
 
+    dists = df.groupby("person").agg(dist=("beelineDist", "sum"))
+
+    # Trips weighted by distance during the whole day
+    dist_weight = df.beelineDist / dists.loc[df.person].dist.to_numpy()
+    df["dist_weight"] = dist_weight
+
     return TripChoice(df, modes, varying, read_global_income(input_file))