refactor the benefit cost code to be gooder hahaha

no really just put in functions and a main and stuff

scripts/summarize/benefit_cost.py

@@ -27,77 +27,74 @@
 guidefile = model_dir + guidefile
 report_output_location = model_dir + report_output_location
 
+
+def read_in_data(file_name, data_names):
+    my_store = h5py.File(file_name, "r+")
+    trips_base = my_store['Trip']
+    var_dict = {}
 
-#read in the data
-my_store_base = h5py.File(h5_base_file, "r+")
-my_store_scen = h5py.File(h5_scen_file, "r+")
-
-trips_base = my_store_base['Trip']
-trips_scen = my_store_scen['Trip']
-
-otaz_base = np.asarray(trips_base["otaz"])
-otaz_base = otaz_base.astype('int')
+    for name in data_names:
+        var_dict[name]= np.asarray(trips_base[name])
 
-dtaz_base = np.asarray(trips_base["dtaz"])
-dtaz_base = dtaz_base.astype('int')
+    return var_dict
 
-#vot_base = np.asarray(trips_base["vot"])
-time_base = np.asarray(trips_base["travtime"])
-#cost_base =vot_base*time_base
+def fill_trip_matrices(output):
 
-otaz_scen = np.asarray(trips_scen["otaz"])
-otaz_scen = otaz_scen.astype('int')
+    trips=np.zeros((zone_dim,zone_dim), np.float64)
 
-dtaz_scen = np.asarray(trips_scen["dtaz"])
-dtaz_scen = dtaz_scen.astype('int')
+    for trip in range (0, len(output['otaz'])-1):
+        this_otaz =  output['otaz'][trip]
+        this_dtaz = output['dtaz'][trip]
+        trips[this_otaz, this_dtaz] =  trips[this_otaz, this_dtaz] + 1
+
+    return trips
 
-#vot_scen = np.asarray(trips_scen["vot"])
-time_scen = np.asarray(trips_scen["travtime"])
-#cost_scen =vot_scen*time_scen
 
+def fill_time_matrices(output):
+    travel_times =np.zeros((zone_dim,zone_dim), np.float64)
 
-trip_base_matrix = np.zeros((zone_dim,zone_dim), np.float64)
-time_base_matrix= np.zeros((zone_dim,zone_dim), np.float64)
-trip_scen_matrix = np.zeros((zone_dim,zone_dim), np.float64)
-time_scen_matrix= np.zeros((zone_dim,zone_dim), np.float64)
+    for trip in range (0, len(output['otaz'])-1):
+        this_otaz =  output['otaz'][trip]
+        this_dtaz = output['dtaz'][trip]
+        this_time = output['travtime'][trip]
+        travel_times[this_otaz, this_dtaz] =  travel_times[this_otaz, this_dtaz] + this_time
 
-# Fill up the base matrices from the arrays of data because they will be faster to work with
-for trip in range (0, len(otaz_base)-1):
-    this_otaz = otaz_base[trip]
-    this_dtaz = dtaz_base[trip]
-    this_time = time_base[trip]
-    trip_base_matrix [this_otaz, this_dtaz] = trip_base_matrix[this_otaz, this_dtaz] + 1
-    time_base_matrix [this_otaz, this_dtaz] = time_base_matrix[this_otaz, this_dtaz] + this_time
+    return travel_times
 
-# Fill up the scenario matrices
-for trip_s in range (0, len(otaz_scen)-1):
-    this_otaz = otaz_scen[trip_s]
-    this_dtaz = dtaz_scen[trip_s]
-    this_time = time_scen[trip_s]
-    trip_scen_matrix [this_otaz, this_dtaz] = trip_scen_matrix[this_otaz, this_dtaz] + 1
-    time_scen_matrix [this_otaz, this_dtaz] = time_scen_matrix[this_otaz, this_dtaz] + this_time
+def calculate_diffs(base_trips, base_travel_time, scen_trips, scen_travel_time):
+    time_diff = (scen_travel_time - base_travel_time)
+    trips_total = (scen_trips + base_trips)
 
-time_diff = (time_scen_matrix - time_base_matrix)
-trips_total = (trip_scen_matrix + trip_base_matrix)
+    avg_time_diff_matrix = time_diff/trips_total
+    avg_time_diff_matrix[np.isnan(avg_time_diff_matrix)] =0
+    return avg_time_diff_matrix
 
-avg_time_diff_matrix = time_diff/trips_total
-avg_time_diff_matrix[np.isnan(avg_time_diff_matrix)] =0
 
-taz_labels = np.array(range(1, zone_dim + 1))
+def write_results(avg_time_diff_matrix):
+    taz_labels = np.array(range(1, zone_dim + 1))
+    time_by_origin = np.mean(avg_time_diff_matrix, axis = 1, dtype=np.float64)
+    time_by_destination = np.mean(avg_time_diff_matrix, axis = 0, dtype=np.float64)
+    all_colls = zip(taz_labels,time_by_origin, time_by_destination)
+    results = pd.DataFrame(data=all_colls, columns=['taz','origin_time_diff', 'destination_time_diff'])
+    results.to_csv(report_output_location)
 
-#benefits by origin:
-time_by_origin = np.mean(avg_time_diff_matrix, axis = 1, dtype=np.float64)
-
-#benefits by destination:
-time_by_destination = np.mean(avg_time_diff_matrix, axis = 0, dtype=np.float64)
-
-#combine arrays:
-all_colls = zip(taz_labels,time_by_origin, time_by_destination)
-
-#convert to dataframe
-results = pd.DataFrame(data=all_colls, columns=['taz','origin_time_diff', 'destination_time_diff'])
+def main():
+    variables_to_read =['otaz', 'dtaz', 'travtime']
+    matrices_to_fill = ['trips', 'travtime']
+    outputs_base = {}
+    outputs_scenario = {}
+    matrices_base = {}
+    matrices_scenario = {}
 
-results.to_csv(report_output_location)
-
+    outputs_base =read_in_data(h5_base_file,variables_to_read)
+    outputs_scenario = read_in_data(h5_scen_file,variables_to_read)
+    base_trips = fill_trip_matrices(outputs_base)
+    base_trav_times = fill_time_matrices(outputs_base)
+    scen_trips = fill_trip_matrices(outputs_scen)
+    scen_trav_times = fill_time_matrices(outputs_scen)
 
+    time_diff_matrix = calculate_diffs(base_trips, base_travel_times, scen_trips, scen_trav_time)
+    write_results(time_diff_matrix)
 
+if __name__ == "__main__":
+    main()