addresses code review

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])

.github/workflows/core_tests.yml

@@ -130,6 +130,7 @@ jobs:
 
  - run: uv run pytest test/test_skim_name_conflicts.py
  - run: uv run pytest test/random_seed/test_random_seed.py
+ - run: uv run pytest test/skip_failed_choices/test_skip_failed_choices.py
  - run: uv run pytest test/trace_id/test_trace_id.py
 
  builtin_regional_models:
@@ -141,11 +142,8 @@ jobs:
  matrix:
  region:
  - prototype_mtc
- - prototype_arc
  - placeholder_psrc
- - prototype_marin
  - prototype_mtc_extended
- - prototype_sandag_xborder
  - production_semcog
  - prototype_mwcog
  - placeholder_multiple_zone

.pre-commit-config.yaml

@@ -17,7 +17,7 @@ repos:
  rev: 22.6.0
  hooks:
  - id: black
- language_version: python3.13
+ language_version: python3.12
  # explicitly set the Python version for the hook environment for black,
  # as older versions of black don't work with the latest Python.
 

activitysim/abm/models/__init__.py

@@ -13,7 +13,6 @@
  disaggregate_accessibility,
  free_parking,
  initialize,
- initialize_los,
  initialize_tours,
  input_checker,
  joint_tour_composition,

activitysim/abm/models/accessibility.py

@@ -154,9 +154,6 @@ def compute_accessibilities_for_zones(
  locals_d["skim_od"] = skim_dict.wrap("orig", "dest").set_df(od_df)
  locals_d["skim_do"] = skim_dict.wrap("dest", "orig").set_df(od_df)
 
- if network_los.zone_system == los.THREE_ZONE:
- locals_d["tvpb"] = network_los.tvpb
-
  logger.info(f"{trace_label}: assign.assign_variables")
  results, trace_results, trace_assigned_locals = assign.assign_variables(
  state,

activitysim/abm/models/atwork_subtour_mode_choice.py

@@ -97,38 +97,6 @@ def atwork_subtour_mode_choice(
  "in_time_col_name": in_time_col_name,
  }
 
- if network_los.zone_system == los.THREE_ZONE:
- # fixme - is this a lightweight object?
- tvpb = network_los.tvpb
-
- tvpb_logsum_odt = tvpb.wrap_logsum(
- orig_key=orig_col_name,
- dest_key=dest_col_name,
- tod_key="out_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_odt",
- )
- tvpb_logsum_dot = tvpb.wrap_logsum(
- orig_key=dest_col_name,
- dest_key=orig_col_name,
- tod_key="in_period",
- segment_key="demographic_segment",
- cache_choices=True,
- trace_label=trace_label,
- tag="tvpb_logsum_dot",
- )
-
- skims.update(
- {"tvpb_logsum_odt": tvpb_logsum_odt, "tvpb_logsum_dot": tvpb_logsum_dot}
- )
-
- # TVPB constants can appear in expressions
- constants.update(
- network_los.setting("TVPB_SETTINGS.tour_mode_choice.CONSTANTS")
- )
-
  estimator = estimation.manager.begin_estimation(state, "atwork_subtour_mode_choice")
  if estimator:
  estimator.write_coefficients(model_settings=model_settings)
@@ -152,32 +120,6 @@ def atwork_subtour_mode_choice(
  trace_choice_name="tour_mode_choice",
  )
 
- # add cached tvpb_logsum tap choices for modes specified in tvpb_mode_path_types
- if network_los.zone_system == los.THREE_ZONE:
- tvpb_mode_path_types = model_settings.tvpb_mode_path_types
- for mode, path_types in tvpb_mode_path_types.items():
- for direction, skim in zip(
- ["od", "do"], [tvpb_logsum_odt, tvpb_logsum_dot]
- ):
- path_type = path_types[direction]
- skim_cache = skim.cache[path_type]
-
- print(f"mode {mode} direction {direction} path_type {path_type}")
-
- for c in skim_cache:
- dest_col = f"{direction}_{c}"
-
- if dest_col not in choices_df:
- choices_df[dest_col] = (
- np.nan
- if pd.api.types.is_numeric_dtype(skim_cache[c])
- else ""
- )
-
- choices_df[dest_col].where(
- choices_df.tour_mode != mode, skim_cache[c], inplace=True
- )
-
  if estimator:
  estimator.write_choices(choices_df[mode_column_name])
  choices_df[mode_column_name] = estimator.get_survey_values(

activitysim/abm/models/disaggregate_accessibility.py

@@ -22,7 +22,6 @@
 )
 from activitysim.core.configuration.logit import TourLocationComponentSettings
 from activitysim.core.expressions import assign_columns
-from activitysim.core.interaction_sample import _resolve_sample_method
 
 logger = logging.getLogger(__name__)
 
@@ -654,6 +653,8 @@ def create_proto_pop(self):
  households.name, persons.name, tours.name = klist
 
  # Create hhid
+ # the households dataframe created by generate_replicates above is indexed by a
+ # simple zero-based RangeIndex, so we will create new ID's by adding 1.
  households[hhid] = households.index + 1
  households["household_serial_no"] = households[hhid]
 
@@ -665,8 +666,14 @@ def create_proto_pop(self):
  .set_index("index")
  .rename(columns={"hhid": hhid})
  )
+ # NOTE: in order to get a stable and reproducible sort of persons here,
+ # the sort keys need to be given all in a single sort command. This code
+ # originally sorted the join (to in theory get all the persons joined in
+ # order, then sorted again by hhid to get the households order, but this
+ # does not guarantee that the persons stay in order, and sorting implementations
+ # can and do vary by platform and dependency version.
  persons = (
- rep.join(persons, sort=True).sort_values(hhid).reset_index(drop=True)
+ rep.join(persons).sort_values([hhid, "index"]).reset_index(drop=True)
  )
  persons[perid] = persons.index + 1
 
@@ -675,6 +682,10 @@ def create_proto_pop(self):
  tours = tours.merge(
  persons[[pkey, hhid, perid]], left_on=tkey, right_on=pkey
  )
+ # We sort tours on the three keys, then drop the index (which is just a row number but scrambled),
+ # then set the tour id, based on the sorted row number. This is to ensure that the tour ids are
+ # assigned in a stable way that is not dependent on the order of the merge.
+ tours = tours.sort_values([hhid, perid, tkey]).reset_index(drop=True)
  tours.index = tours.index.set_names([tourid])
  tours.index += 1
  tours = tours.reset_index().drop(columns=[pkey])