PNUM hard-coded and Vehicle table creation bug fix (#953)

* minor fixes for SimOR development

* blacken

* Using PNUM if available in JTFC

Required to maintain backwards compatibility in tour indexes in SANDAG models

---------

Co-authored-by: Jeffrey Newman <[email protected]>

Author: dhensle

activitysim/abm/models/joint_tour_frequency_composition.py

@@ -146,8 +146,15 @@ def joint_tour_frequency_composition(
     # - but we don't know the tour participants yet
     # - so we arbitrarily choose the first person in the household
     # - to be point person for the purpose of generating an index and setting origin
-    # FIXME: not all models are guaranteed to have PNUM
-    temp_point_persons = persons.loc[persons.PNUM == 1]
+    if "PNUM" in persons.columns:
+        temp_point_persons = persons.loc[persons.PNUM == 1]
+    else:
+        # if PNUM is not available, we can still get the first person in the household
+        temp_point_persons = (
+            persons.sort_index()  # ensure stable ordering
+            .groupby("household_id", as_index=False)
+            .first()
+        )
     temp_point_persons["person_id"] = temp_point_persons.index
     temp_point_persons = temp_point_persons.set_index("household_id")
     temp_point_persons = temp_point_persons[["person_id", "home_zone_id"]]

activitysim/abm/models/joint_tour_participation.py

@@ -60,6 +60,9 @@ def joint_tour_participation_candidates(joint_tours, persons_merged):
     # if this happens, participant_id may not be unique
     # channel random seeds will overlap at MAX_PARTICIPANT_PNUM (probably not a big deal)
     # and estimation infer will fail
+    if "PNUM" not in candidates.columns:
+        # create a PNUM column that just numbers the candidates for assignment of participant_id
+        candidates["PNUM"] = candidates.groupby("household_id").cumcount() + 1
     assert (
         candidates.PNUM.max() < MAX_PARTICIPANT_PNUM
     ), f"max persons.PNUM ({candidates.PNUM.max()}) > MAX_PARTICIPANT_PNUM ({MAX_PARTICIPANT_PNUM})"

activitysim/abm/tables/vehicles.py

@@ -29,6 +29,12 @@ def vehicles(state: workflow.State, households: pd.DataFrame):
     """
 
     # initialize vehicles table
+    if "auto_ownership" not in households.columns:
+        # grab the proto_households table instead
+        # this is called when running disaggregate accessibilities and the vehicles table is used in the logsum calculation
+        households = state.get_table("proto_households")
+        households.index.name = "household_id"
+
     vehicles = households.loc[households.index.repeat(households["auto_ownership"])]
     vehicles = vehicles.reset_index()[["household_id"]]