cdap choose backfill persons randomly like ctramp

activitysim/abm/models/util/cdap.py

@@ -14,6 +14,7 @@
 from zbox import toolz as tz, gen
 
 from activitysim.core import simulate
+from activitysim.core import pipeline
 
 from activitysim.core import chunk
 from activitysim.core import logit
@@ -90,7 +91,7 @@ def assign_cdap_rank(persons, trace_hh_id=None, trace_label=None):
     than 5 people, the cdapPersonArray is the same as the person array."
 
     We diverge from the above description in that a cdap_rank is assigned to all persons,
-    including 'extra' householld members, whose activity is assigned subsequently.
+    including 'extra' household members, whose activity is assigned subsequently.
     The pair _hh_id_, cdap_rank will uniquely identify each household member.
 
     Parameters
@@ -130,10 +131,19 @@ def assign_cdap_rank(persons, trace_hh_id=None, trace_label=None):
     del children
 
     # choose up to MAX_HHSIZE, preferring anyone already chosen
+    # others = \
+    #     persons[[_hh_id_, 'cdap_rank']]\
+    #     .sort_values(by=[_hh_id_, 'cdap_rank'], ascending=[True, True])\
+    #     .groupby(_hh_id_).head(MAX_HHSIZE)
+
+    # choose up to MAX_HHSIZE, choosing randomly
+    others = persons[[_hh_id_, 'cdap_rank']].copy()
+    others['random_order'] = pipeline.get_rn_generator().random_for_df(persons)
     others = \
-        persons[[_hh_id_, 'cdap_rank']]\
-        .sort_values(by=[_hh_id_, 'cdap_rank'], ascending=[True, True])\
+        others\
+        .sort_values(by=[_hh_id_, 'random_order'], ascending=[True, True])\
         .groupby(_hh_id_).head(MAX_HHSIZE)
+
     # tag the backfilled persons
     persons.loc[others[others.cdap_rank == RANK_UNASSIGNED].index, 'cdap_rank'] \
         = RANK_BACKFILL

activitysim/abm/test/configs/cdap_indiv_and_hhsize1.csv

@@ -9,10 +9,8 @@ Pre-driving-age child who is in school alternative-specific constants,ptype == 7
 Pre-driving-age child who is in school interaction with age 6 to 9,(ptype == 7) & (age >= 6) & (age <= 9),-0.2943,,
 Pre-driving-age child who is in school interaction with age 13 to 15,(ptype == 7) & (age >= 13) & (age <= 15),-0.7141,-0.672,
 Pre-driving-age child who is too young for school alternative-specific constants,ptype == 8,1.052531189,-0.837567776,
-# tm1 age bug,,,,
-Pre-driving-age child who is too young for school interaction with age 0 to 1,(ptype == 8) * age,-0.4515,,
-# asim corrected age bug,,,,
-# Pre-driving-age child who is too young for school interaction with age 0 to 1,(ptype == 8) & (age >= 0) & (age <= 1),-0.4515,,
+# corrected tm1 age bug,,,,
+Pre-driving-age child who is too young for school interaction with age 0 to 1,(ptype == 8) & (age >= 0) & (age <= 1),-0.4515,,
 Pre-driving-age child who is too young for school interaction with age 4 to 5,(ptype == 8) & (age >= 4) & (age <= 5),0.6107,,
 #,,,,
 Full-time worker interaction with age less than 40,(ptype == 1) & (age < 40),0.2091,,

activitysim/abm/test/test_pipeline.py

@@ -23,9 +23,9 @@
 HOUSEHOLDS_SAMPLE_SIZE = 100
 
 # household with mandatory, non mandatory, atwork_subtours, and joint tours
-HH_ID = 1396417
+HH_ID = 1402945
 
-#  [1081630 1396417 1511245 1594943 1747572 1931915 2222690 2366390 2727112]
+# [1402945 1511245 1931827 1931908 2224336 2307195 2366390 2408855 2727112 1196229  932096  932412]
 
 # SKIP_FULL_RUN = True
 SKIP_FULL_RUN = False
@@ -318,7 +318,7 @@ def get_trace_csv(file_name):
     return df
 
 
-EXPECT_TOUR_COUNT = 210
+EXPECT_TOUR_COUNT = 205
 
 
 def regress_tour_modes(tours_df):
@@ -334,54 +334,46 @@ def regress_tour_modes(tours_df):
     """
                  tour_mode  person_id tour_type  tour_num  tour_category
     tour_id
-    91533592         WALK    3156330  othmaint         1  non_mandatory
-    91533620     WALK_LOC    3156331  othdiscr         1  non_mandatory
-    91533647  SHARED2FREE    3156332     maint         1         atwork
-    91533655         WALK    3156332      work         1      mandatory
-    91533650     WALK_LOC    3156332  othmaint         1  non_mandatory
-    91533583  SHARED2FREE    3156333  othmaint         1          joint
-    91533680         WALK    3156333    school         1      mandatory
-    91533683  SHARED2FREE    3156333    social         1  non_mandatory
-    91533709         WALK    3156334    school         1      mandatory
-    91533738         WALK    3156335    school         1      mandatory
+    92466005         WALK    3188482      work         1      mandatory
+    92465999         WALK    3188482  othdiscr         1  non_mandatory
+    92466034     WALK_LOC    3188483      work         1      mandatory
+    92466040         WALK    3188484       eat         1         atwork
+    92465989  SHARED2FREE    3188484  othdiscr         1          joint
+    92466063     WALK_LOC    3188484      work         1      mandatory
+    92466059     WALK_LOC    3188484    school         2      mandatory
+    92466092         WALK    3188485      work         1      mandatory
     """
 
     EXPECT_PERSON_IDS = [
-        3156330,
-        3156331,
-        3156332,
-        3156332,
-        3156332,
-        3156333,
-        3156333,
-        3156333,
-        3156334,
-        3156335,
+        3188482,
+        3188482,
+        3188483,
+        3188484,
+        3188484,
+        3188484,
+        3188484,
+        3188485,
         ]
 
     EXPECT_TOUR_TYPES = [
-        'othmaint',
+        'work',
+        'othdiscr',
+        'work',
+        'eat',
         'othdiscr',
-        'maint',
         'work',
-        'othmaint',
-        'othmaint',
-        'school',
-        'social',
-        'school',
         'school',
+        'work',
         ]
 
     EXPECT_MODES = [
         'WALK',
-        'WALK_LOC',
-        'SHARED2FREE',
         'WALK',
         'WALK_LOC',
-        'SHARED2FREE',
         'WALK',
         'SHARED2FREE',
-        'WALK',
+        'WALK_LOC',
+        'WALK_LOC',
         'WALK',
         ]