Route identification

gtfs_funnel/Makefile

@@ -14,6 +14,11 @@ preprocess:
 	python cleanup.py
 	python route_typologies.py
 
+# Start compiling scripts needed when we start concatenating several days
+# Clean route names for displaying across time
+timeseries_preprocessing: 
+	python clean_route_naming.py
+
 
 funnel_gtfs_data: 
 	make download_gtfs_data && make preprocess

gtfs_funnel/clean_route_naming.py

@@ -0,0 +1,77 @@
+"""
+Create file with cleaned up route info.
+"""
+import pandas as pd
+
+from segment_speed_utils import gtfs_schedule_wrangling, helpers
+from segment_speed_utils.project_vars import SCHED_GCS
+from shared_utils import rt_dates
+
+def concatenate_routes_across_dates(analysis_date_list: list) -> pd.DataFrame:
+    """
+    """
+    df = pd.concat([
+        helpers.import_scheduled_trips(
+            analysis_date,
+            columns = ["gtfs_dataset_key", "name", "route_id", 
+                       "route_long_name", "route_short_name", "route_desc"],
+            get_pandas = True
+        ).assign(
+            service_date = pd.to_datetime(analysis_date)
+        ) for analysis_date in analysis_date_list
+    ], axis=0, ignore_index=True)
+
+    # Fill in missing values
+    df = df.assign(
+        route_id = df.route_id.fillna(""),
+        route_short_name = df.route_short_name.fillna(""),
+        route_long_name = df.route_long_name.fillna(""),
+    )
+
+    df = df.assign(
+        combined_name = df.route_short_name + "__" + df.route_long_name
+    )
+
+    return df
+
+if __name__ == "__main__":
+
+    from update_vars import CONFIG_DICT
+    CLEANED_ROUTE_NAMING = CONFIG_DICT["route_identification_file"]
+
+    months = [
+        "mar", "apr", "may",
+        "jun", "jul", "aug",
+        "sep", "oct", "nov",
+        "dec"
+    ]
+
+    date_list = [
+        rt_dates.DATES[f"{m}2023"] 
+        for m in months
+    ]
+
+    df =  concatenate_routes_across_dates(date_list)
+
+    df = df.assign(
+        route_id2 = df.apply(
+            lambda x: 
+            gtfs_schedule_wrangling.standardize_route_id(
+                x, "name", "route_id"), 
+            axis=1)
+    )
+
+    route_cols = ["schedule_gtfs_dataset_key", "name", "route_id2"]
+
+    df2 = gtfs_schedule_wrangling.most_recent_route_info(
+        df,
+        group_cols = route_cols,
+        route_col = "combined_name"
+    ).pipe(
+        gtfs_schedule_wrangling.most_recent_route_info, 
+        group_cols = ["schedule_gtfs_dataset_key", "name", 
+                      "recent_combined_name"],
+        route_col = "route_id2"
+    )
+
+    df2.to_parquet(f"{SCHED_GCS}{CLEANED_ROUTE_NAMING}.parquet")

gtfs_funnel/config.yml

@@ -4,4 +4,5 @@ timestamp_col: "location_timestamp_local"
 time_min_cutoff: 10
 stop_times_direction_file: "stop_times_direction"
 trip_metrics_file: "schedule_trip_metrics"
-route_direction_metrics_file: "schedule_route_direction_metrics"
+route_direction_metrics_file: "schedule_route_direction_metrics"
+route_identification_file: "standardized_route_ids"