get categories from alternatives

wsp-sag · Dec 12, 2023 · e1e5d41 · e1e5d41
1 parent 51f9a24
commit e1e5d41
Show file tree

Hide file tree

Showing 11 changed files with 57 additions and 70 deletions.
diff --git a/activitysim/abm/models/atwork_subtour_frequency.py b/activitysim/abm/models/atwork_subtour_frequency.py
@@ -23,7 +23,7 @@
 def add_null_results(state, trace_label, tours):
     logger.info("Skipping %s: add_null_results", trace_label)
     cat_type = pd.api.types.CategoricalDtype(
-        ["", "no_subtours", "eat", "business1", "maint", "business2", "eat_business"],
+        [""],
         ordered=False,
     )
     choices = choices.astype(cat_type)
@@ -109,7 +109,7 @@ def atwork_subtour_frequency(
     # convert indexes to alternative names
     choices = pd.Series(model_spec.columns[choices.values], index=choices.index)
     cat_type = pd.api.types.CategoricalDtype(
-        ["", "no_subtours", "eat", "business1", "maint", "business2", "eat_business"],
+        alternatives.index.tolist()+[""],
         ordered=False,
     )
     choices = choices.astype(cat_type)
@@ -133,6 +133,12 @@ def atwork_subtour_frequency(
 
     subtours = process_atwork_subtours(state, work_tours, alternatives)
 
+    # convert purpose to pandas categoricals
+    purpose_type = pd.api.types.CategoricalDtype(
+        alternatives.columns.tolist()+["atwork"], ordered=False
+    )
+    subtours["tour_type"] = subtours["tour_type"].astype(purpose_type)
+
     tours = state.extend_table("tours", subtours)
 
     state.tracing.register_traceable_table("tours", subtours)

diff --git a/activitysim/abm/models/joint_tour_composition.py b/activitysim/abm/models/joint_tour_composition.py
@@ -111,7 +111,7 @@ def joint_tour_composition(
     # convert indexes to alternative names
     choices = pd.Series(model_spec.columns[choices.values], index=choices.index)
     cat_type = pd.api.types.CategoricalDtype(
-        ["", "adults", "children", "mixed"], ordered=False
+        model_spec.columns.tolist()+[""], ordered=False
     )
     choices = choices.astype(cat_type)
 

diff --git a/activitysim/abm/models/joint_tour_frequency.py b/activitysim/abm/models/joint_tour_frequency.py
@@ -99,29 +99,7 @@ def joint_tour_frequency(
     # convert indexes to alternative names
     choices = pd.Series(model_spec.columns[choices.values], index=choices.index)
     cat_type = pd.api.types.CategoricalDtype(
-        [
-            "0_tours",
-            "1_Shop",
-            "1_Main",
-            "1_Eat",
-            "1_Visit",
-            "1_Disc",
-            "2_SS",
-            "2_SM",
-            "2_SE",
-            "2_SV",
-            "2_SD",
-            "2_MM",
-            "2_ME",
-            "2_MV",
-            "2_MD",
-            "2_EE",
-            "2_EV",
-            "2_ED",
-            "2_VV",
-            "2_VD",
-            "2_DD",
-        ],
+        model_spec.columns.tolist(),
         ordered=False,
     )
     choices = choices.astype(cat_type)
@@ -147,6 +125,12 @@ def joint_tour_frequency(
 
     joint_tours = process_joint_tours(state, choices, alternatives, temp_point_persons)
 
+    # convert purpose to pandas categoricals
+    purpose_type = pd.api.types.CategoricalDtype(
+        alternatives.columns.tolist(), ordered=False
+    )
+    joint_tours["tour_type"] = joint_tours["tour_type"].astype(purpose_type)
+
     tours = state.extend_table("tours", joint_tours)
 
     state.tracing.register_traceable_table("tours", joint_tours)

diff --git a/activitysim/abm/models/mandatory_tour_frequency.py b/activitysim/abm/models/mandatory_tour_frequency.py
@@ -115,7 +115,7 @@ def mandatory_tour_frequency(
     # convert indexes to alternative names
     choices = pd.Series(model_spec.columns[choices.values], index=choices.index)
     cat_type = pd.api.types.CategoricalDtype(
-        ["", "work1", "work2", "school1", "school2", "work_and_school"], ordered=False
+        model_spec.columns.tolist()+[""], ordered=False
     )
     choices = choices.astype(cat_type)
 
@@ -142,6 +142,12 @@ def mandatory_tour_frequency(
         state, persons=choosers, mandatory_tour_frequency_alts=alternatives
     )
 
+    # convert purpose to pandas categoricals
+    purpose_type = pd.api.types.CategoricalDtype(
+        alternatives.columns.tolist()+["univ","home","escort"], ordered=False
+    )
+    mandatory_tours["tour_type"] = mandatory_tours["tour_type"].astype(purpose_type)
+
     tours = state.extend_table("tours", mandatory_tours)
     state.tracing.register_traceable_table("tours", mandatory_tours)
     state.get_rn_generator().add_channel("tours", mandatory_tours)

diff --git a/activitysim/abm/models/non_mandatory_tour_frequency.py b/activitysim/abm/models/non_mandatory_tour_frequency.py
@@ -354,6 +354,12 @@ def non_mandatory_tour_frequency(
     )
     assert len(non_mandatory_tours) == extended_tour_counts.sum().sum()
 
+    # convert purpose to pandas categoricals
+    purpose_type = pd.api.types.CategoricalDtype(
+        alternatives.columns.tolist(), ordered=False
+    )
+    non_mandatory_tours["tour_type"] = non_mandatory_tours["tour_type"].astype(purpose_type)
+
     if estimator:
         # make sure they created the right tours
         survey_tours = estimation.manager.get_survey_table("tours").sort_index()

diff --git a/activitysim/abm/models/telecommute_frequency.py b/activitysim/abm/models/telecommute_frequency.py
@@ -87,7 +87,7 @@ def telecommute_frequency(
 
     choices = pd.Series(model_spec.columns[choices.values], index=choices.index)
     telecommute_frequency_cat = pd.api.types.CategoricalDtype(
-        ["", "No_Telecommute", "1_day_week", "2_3_days_week", "4_days_week"],
+        model_spec.columns.tolist()+[""],
         ordered=False,
     )
     choices = choices.astype(telecommute_frequency_cat)

diff --git a/activitysim/abm/models/tour_mode_choice.py b/activitysim/abm/models/tour_mode_choice.py
@@ -68,10 +68,10 @@ def create_logsum_trips(
     pandas.DataFrame
         Table of trips: 2 per tour, with O/D and purpose inherited from tour
     """
+    stop_frequency_alts = state.get_injectable("stop_frequency_alts")
     stop_freq_cat_type = pd.api.types.CategoricalDtype(
-        ["", "work1", "work2", "school1", "school2", "work_and_school"], ordered=False
+        stop_frequency_alts.index.tolist()+[""], ordered=False
     )
-    stop_frequency_alts = state.get_injectable("stop_frequency_alts")
     stop_freq = "0out_0in"  # no intermediate stops
     tours["stop_frequency"] = stop_freq
     tours["stop_frequency"] = tours["stop_frequency"].astype(stop_freq_cat_type)

diff --git a/activitysim/abm/models/trip_purpose.py b/activitysim/abm/models/trip_purpose.py
@@ -163,6 +163,12 @@ def choose_intermediate_trip_purpose(
         state.tracing.trace_df(rands, "%s.rands" % trace_label, columns=[None, "rand"])
 
     choices = choices.map(pd.Series(purpose_cols))
+    # expand the purpose categorical
+    for p in purpose_cols:
+        if not p in trips.primary_purpose.cat.categories:
+            trips.primary_purpose = trips.primary_purpose.cat.add_categories(
+                [p]
+            )
     choices = choices.astype(trips["primary_purpose"].dtype)
     return choices
 
@@ -208,6 +214,13 @@ def run_trip_purpose(state: workflow.State, trips_df, estimator, trace_label):
 
     result_list = []
 
+    # add home to purpose categorical
+    # check if parking_name is in the purpose category
+    if not "home" in trips_df.primary_purpose.cat.categories:
+        trips_df.primary_purpose = trips_df.primary_purpose.cat.add_categories(
+            ["home"]
+        )
+
     # - last trip of outbound tour gets primary_purpose
     last_trip = trips_df.trip_num == trips_df.trip_count
     purpose = trips_df.primary_purpose[last_trip & trips_df.outbound]

diff --git a/activitysim/abm/models/util/tour_frequency.py b/activitysim/abm/models/util/tour_frequency.py
@@ -57,26 +57,6 @@ def create_tours(tour_counts, tour_category, parent_col="person_id"):
     # reformat with the columns given below
     tours = tour_counts.stack().reset_index()
     tours.columns = [parent_col, "tour_type", "tour_type_count"]
-    cat_tour_type = pd.api.types.CategoricalDtype(
-        [
-            "work",
-            "school",
-            "univ",
-            "escort",
-            "eatout",
-            "shopping",
-            "social",
-            "othmaint",
-            "othdiscr",
-            "eat",
-            "business",
-            "maint",
-            "atwork",
-            "home",
-        ],
-        ordered=False,
-    )
-    tours["tour_type"] = tours["tour_type"].astype(cat_tour_type)
 
     """
         <parent_col> tour_type  tour_type_count
@@ -724,26 +704,6 @@ def create_joint_tours(
     tours_purp.columns = [parent_col, "tour_id_temp", "tour_type"]
     tours_purp["tour_id_temp"] = range(1, 1 + len(tours_purp))
     tours_purp["tour_type"] = tours_purp["tour_type"].map(tour_type_dict)
-    cat_tour_type = pd.api.types.CategoricalDtype(
-        [
-            "work",
-            "school",
-            "univ",
-            "escort",
-            "eatout",
-            "shopping",
-            "social",
-            "othmaint",
-            "othdiscr",
-            "eat",
-            "business",
-            "maint",
-            "atwork",
-            "home",
-        ],
-        ordered=False,
-    )
-    tours_purp["tour_type"] = tours_purp["tour_type"].astype(cat_tour_type)
 
     """
         <parent_col> tour_id_temp  tour_type

diff --git a/activitysim/core/workflow/state.py b/activitysim/core/workflow/state.py
@@ -1080,6 +1080,16 @@ def extend_table(self, table_name, df, axis=0):
                 new_df_columns = [c for c in df.columns if c not in table_df.columns]
                 df = df[new_df_columns]
                 missing_df_str_columns = []
+
+            # union categoricals
+            for c in table_df.columns.intersection(df.columns):
+                if isinstance(table_df[c].dtype,pd.api.types.CategoricalDtype):
+                    if isinstance(df[c].dtype,pd.api.types.CategoricalDtype):
+                        from pandas.api.types import union_categoricals
+
+                        uc = union_categoricals([table_df[c], df[c]])
+                        table_df[c]=pd.Categorical(table_df[c],categories=uc.categories)
+                        df[c]=pd.Categorical(df[c],categories=uc.categories)
 
             # preserve existing column order
             df = pd.concat([table_df, df], sort=False, axis=axis)

diff --git a/...im/examples/prototype_mtc_extended/configs/stop_frequency_annotate_tours_preprocessor.csv b/...im/examples/prototype_mtc_extended/configs/stop_frequency_annotate_tours_preprocessor.csv
@@ -47,5 +47,7 @@ AccesibilityAtDestination if transit,pracc,"pracc.where(~tour_mode_is_transit, _
 AccesibilityAtDestination if non_motorized,pracc,"pracc.where(~tour_mode_is_non_motorized, reindex(accessibility.nmRetail, df.destination))"
 ,destination_area_type,"reindex(land_use.area_type, df.destination)"
 ,is_mixed,df.composition=='mixed'
+,is_adults_tour,df.composition=='adults'
+,is_social,primary_purpose=='social'
 ,is_outbound_school_escort,"(df.school_esc_outbound.isin(['ride_share', 'pure_escort']))"
 ,is_inbound_school_escort,"(df.school_esc_inbound.isin(['ride_share', 'pure_escort']))"