Merge pull request #1319 from cal-itp/remove-vp-nn-output

Remove vp nn output
cal-itp · Dec 9, 2024 · 323bbed · 323bbed
2 parents cbac204 + ebd0676
commit 323bbed
Show file tree

Hide file tree

Showing 10 changed files with 575 additions and 161 deletions.
diff --git a/_shared_utils/requirements.txt b/_shared_utils/requirements.txt
@@ -1,7 +1,7 @@
 -e .
 altair==5.3.0
 altair-transform==0.2.0
-gtfs-segments==0.1.0
+gtfs-segments==2.1.7
 pyairtable==2.2.2
 great_tables==0.14.0
 omegaconf==2.3.0 # better yaml configuration

diff --git a/_shared_utils/shared_utils/geo_utils.py b/_shared_utils/shared_utils/geo_utils.py
@@ -1,6 +1,8 @@
 """
 Geospatial utility functions
 """
+from typing import Union
+
 import geopandas as gpd
 import numpy as np
 import pandas as pd
@@ -17,13 +19,16 @@
 geo_const_miles = 3_959_000 * np.pi / 180
 
 
-def nearest_snap(line: shapely.LineString, point: shapely.Point, k_neighbors: int = 1) -> np.ndarray:
+def nearest_snap(line: Union[shapely.LineString, np.ndarray], point: shapely.Point, k_neighbors: int = 1) -> np.ndarray:
     """
     Based off of this function,
     but we want to return the index value, rather than the point.
     https://github.com/UTEL-UIUC/gtfs_segments/blob/main/gtfs_segments/geom_utils.py
     """
-    line = np.asarray(line.coords)
+    if isinstance(line, shapely.LineString):
+        line = np.asarray(line.coords)
+    elif isinstance(line, np.ndarray):
+        line = line
     point = np.asarray(point.coords)
     tree = KDTree(line)
 

diff --git a/_shared_utils/shared_utils/gtfs_analytics_data.yml b/_shared_utils/shared_utils/gtfs_analytics_data.yml
@@ -32,7 +32,6 @@ speeds_tables:
   usable_vp: vp_usable
   vp_dwell: vp_usable_dwell
   vp_condensed_line: condensed/vp_condensed
-  vp_nearest_neighbor: condensed/vp_nearest_neighbor
   timestamp_col: ${speed_vars.timestamp_col}
   time_min_cutoff: ${speed_vars.time_min_cutoff}
 

diff --git a/gtfs_funnel/vp_condenser.py b/gtfs_funnel/vp_condenser.py
@@ -58,49 +58,6 @@ def condense_vp_to_linestring(
     return 
 
 
-def prepare_vp_for_all_directions(
-    analysis_date: str, 
-    dict_inputs: dict
-) -> gpd.GeoDataFrame:
-    """
-    For each direction, exclude one the opposite direction and
-    save out the arrays of valid indices.
-    Every trip will have 4 rows, 1 row corresponding to each direction.
-    
-    Ex: for a given trip's northbound points, exclude southbound vp.
-    Subset vp_idx, location_timestamp_local and coordinate arrays 
-    to exclude southbound.
-    """
-    INPUT_FILE = dict_inputs.speeds_tables.vp_condensed_line
-    EXPORT_FILE = dict_inputs.speeds_tables.vp_nearest_neighbor
-
-    vp = delayed(gpd.read_parquet)(
-        f"{SEGMENT_GCS}{INPUT_FILE}_{analysis_date}.parquet",
-    )
-
-    dfs = [
-        delayed(vp_transform.combine_valid_vp_for_direction)(
-            vp, direction) 
-        for direction in vp_transform.ALL_DIRECTIONS
-    ]
-
-    results = [compute(i)[0] for i in dfs]
-
-    gdf = pd.concat(
-        results, axis=0, ignore_index=True
-    ).sort_values(
-        ["trip_instance_key", "vp_primary_direction"]
-    ).reset_index(drop=True)
-
-    utils.geoparquet_gcs_export(
-        gdf,
-        SEGMENT_GCS,
-        f"{EXPORT_FILE}_{analysis_date}"
-    )
-
-    return 
-
-
 if __name__ == "__main__":
 
     from update_vars import analysis_date_list
@@ -116,17 +73,9 @@ def prepare_vp_for_all_directions(
 
         condense_vp_to_linestring(analysis_date, GTFS_DATA_DICT)
 
-        time1 = datetime.datetime.now()
+        end = datetime.datetime.now()
 
         logger.info(
             f"{analysis_date}: condense vp for trip "
-            f"{time1 - start}"
-        )
-
-        prepare_vp_for_all_directions(analysis_date, GTFS_DATA_DICT)
-
-        end = datetime.datetime.now()
-        logger.info(
-            f"{analysis_date}: prepare vp to use in nearest neighbor: "
-            f"{end - time1}"
-        )       
+            f"{end - start}"
+        )