Adding differential sampling

abm/src/main/python/dataExporter/abmScenario.py

@@ -288,6 +288,10 @@ def properties(self) -> dict:
             "rsmSamplingRate":{
                 "line": "rsm.default.sampling.rate=",
                 "type" : "float",
+                "value": None},
+            "useDifferentialSampling":{
+                "line": "use.differential.sampling=",
+                "type" : "int",
                 "value": None}
         }
 
@@ -2777,7 +2781,7 @@ def ie(self) -> pd.DataFrame:
 
         study_area_file = os.path.join(self.scenario_path, "input", "study_area.csv")
 
-        if os.path.exists(study_area_file):
+        if useDifferentialSampling & os.path.exists(study_area_file):
             df = pd.read_csv(study_area_file)
             study_area_taz = set(df['taz'])
             rsm_zone = set(rsm_zones.loc[rsm_zones['taz'].isin(study_area_taz), 'cluster_id'])

abm/src/main/resources/sandag_abm.properties

@@ -1344,6 +1344,7 @@ rsm.zones = 2000
 external.zones = 12
 run.rsm.sampling = 0
 rsm.default.sampling.rate = 0.25
+use.differential.sampling = 1
 rsm.min.sampling.rate = 0.25
 run.rsm.assembler = 0
 rsm.centroid.connector.start.id = 55000

rsm/assembler.py

@@ -34,7 +34,8 @@ def rsm_assemble(
     taz_crosswalk=None,
     sample_rate=0.25,
     study_area_taz=None,
-    run_assembler=1
+    run_assembler=1,
+    differential_sampling=1
 ):
     """
     Assemble and evaluate RSM trip making.
@@ -70,6 +71,7 @@ def rsm_assemble(
     study_area_rsm_zones :  list
         it is list of study area RSM zones
     
+    
     Returns
     -------
     #final_trips_rsm : pd.DataFrame
@@ -239,7 +241,7 @@ def _agg_by_hhid_and_tripmode(df, name):
         final_jnt_trips = scaleup_to_rsm_samplingrate(jnt_trips_rsm, 
                                                       households, 
                                                       taz_crosswalk, 
-                                                      scale_factor, 
+                                                      scale_factor,
                                                       study_area_tazs=sa_rsm) 
 
     return final_ind_trips, final_jnt_trips

rsm/utility.py

@@ -251,7 +251,7 @@ def scaleup_to_rsm_samplingrate(df,
     hh["taz"] = hh["taz"].map(dict_clusters)
     hh['scale_factor'] = scale_factor
 
-    if study_area_tazs:       
+    if study_area_tazs:
         hh.loc[hh['taz'].isin(study_area_tazs), 'scale_factor'] = 1
 
     df = pd.merge(df, hh, left_on='hh_id', right_on='hhid', how='left')

scripts/rsm_assembler.py

@@ -38,7 +38,17 @@
 TAZ_CROSSWALK = os.path.join(rsm_dir, "input", "taz_crosswalk.csv")
 STUDY_AREA = os.path.join(rsm_dir, "input", "study_area.csv")
 
-if os.path.exists(STUDY_AREA):
+#creating copy of individual and joint trips file
+shutil.copy(RSM_INDIV_TRIPS, os.path.join(rsm_dir, "output", "indivTripData_abm_"+ str(iteration) + ".csv"))
+shutil.copy(RSM_JOINT_TRIPS, os.path.join(rsm_dir, "output", "jointTripData_abm_"+ str(iteration) + ".csv"))
+
+ABM_PROPERTIES_FOLDER = os.path.join(rsm_dir, "conf")
+ABM_PROPERTIES = os.path.join(ABM_PROPERTIES_FOLDER, "sandag_abm.properties")
+RUN_ASSEMBLER = int(get_property(ABM_PROPERTIES, "run.rsm.assembler"))
+SAMPLE_RATE = float(get_property(ABM_PROPERTIES, "rsm.default.sampling.rate"))
+USE_DIFFERENTIAL_SAMPLING = int(get_property(ABM_PROPERTIES, "use.differential.sampling"))
+
+if USE_DIFFERENTIAL_SAMPLING & os.path.exists(STUDY_AREA):
     logging.info(f"Study Area file: {STUDY_AREA}")
     study_area_taz = find_rsm_zone_of_study_area(STUDY_AREA, TAZ_CROSSWALK)
     if study_area_taz is not None:
@@ -53,15 +63,6 @@
 else:
     SA_TAZ = None
 
-#creating copy of individual and joint trips file
-shutil.copy(RSM_INDIV_TRIPS, os.path.join(rsm_dir, "output", "indivTripData_abm_"+ str(iteration) + ".csv"))
-shutil.copy(RSM_JOINT_TRIPS, os.path.join(rsm_dir, "output", "jointTripData_abm_"+ str(iteration) + ".csv"))
-
-ABM_PROPERTIES_FOLDER = os.path.join(rsm_dir, "conf")
-ABM_PROPERTIES = os.path.join(ABM_PROPERTIES_FOLDER, "sandag_abm.properties")
-RUN_ASSEMBLER = int(get_property(ABM_PROPERTIES, "run.rsm.assembler"))
-SAMPLE_RATE = float(get_property(ABM_PROPERTIES, "rsm.default.sampling.rate"))
-
 #RSM Assembler
 final_ind_trips, final_jnt_trips = rsm_assemble(
     ORG_INDIV_TRIPS,

scripts/rsm_sampler.py

@@ -47,6 +47,7 @@
 sampling_rate = float(get_property(ABM_PROPERTIES, "rsm.default.sampling.rate"))
 min_sampling_rate = float(get_property(ABM_PROPERTIES, "rsm.min.sampling.rate"))
 baseline_run_dir = get_property(ABM_PROPERTIES, "rsm.baseline.run.dir")
+use_differential_sampling = int(get_property(ABM_PROPERTIES, "use.differential.sampling"))
 
 if run_rsm_sampling == 1:
     CURR_ITER_ACCESS = os.path.join(
@@ -70,7 +71,7 @@
 logging.info(f"Current Iteration Accessibility File: {CURR_ITER_ACCESS}")
 logging.info(f"Previous Iteration Accessibility File: {PREV_ITER_ACCESS}")
 
-if os.path.exists(EXPILICT_AGG_TAZ):
+if use_differential_sampling & os.path.exists(EXPILICT_AGG_TAZ):
     logging.info(f"Study Area file: {EXPILICT_AGG_TAZ}")
     study_area_taz = find_rsm_zone_of_study_area(EXPILICT_AGG_TAZ, OUTPUT_TAZ_CROSSWALK)
     if study_area_taz is not None:
@@ -83,6 +84,7 @@
         sa_taz = None
 
 else:
+    logger.info("All RSM zones will be sampled at the deafult sampling rate")
     sa_taz = None
 
 

scripts/rsm_zone_aggregator.py

@@ -66,6 +66,8 @@
 logging.info("Check if the study area file exists in the RSM input folder")
 if os.path.exists(EXPILICT_AGG_TAZ):
     EXPLICIT_ZONE_AGG = create_list_study_area_taz(EXPILICT_AGG_TAZ)
+    logging.info("The input folder has a study_area file. The TAZs will be aggregated based on the study area file")
+    logging.info(EXPLICIT_ZONE_AGG)
 else:
     EXPLICIT_ZONE_AGG = []