Update API to support predict start coordinates instead SHAPE_Length

Author: dpham2123

Use_Cases/Weather-Aware Routing/Duy Pham/api.py

@@ -1,7 +1,10 @@
+import pandas as pd
+from shapely import wkt
+from shapely.geometry import Point
+import geopandas as gpd
+import joblib
 from fastapi import FastAPI
 from pydantic import BaseModel
-import joblib
-import pandas as pd
 
 model = joblib.load("ev_model.pkl")
 
@@ -14,9 +17,30 @@
     "total_prcp",
 ]
 
-class StationFeatures(BaseModel):
+traffic_df = pd.read_csv("Traffic data.csv")
+
+
+traffic_df["geometry"] = traffic_df["geometry"].apply(wkt.loads)
+
+traffic_gdf = gpd.GeoDataFrame(traffic_df, geometry="geometry", crs="EPSG:4326")
+
+def get_shape_length_from_coords(lat: float, lon: float) -> float:
+    """
+    Given a lat/lon point, find the nearest road segment in traffic_gdf
+    and return its SHAPE_Length.
+    """
+    pt = Point(lon, lat)  
+
+    distances = traffic_gdf.distance(pt)
+    nearest_idx = distances.idxmin()
+
+    shape_length = traffic_gdf.loc[nearest_idx, "SHAPE_Length"]
+    return float(shape_length)
+
+class CoordRequest(BaseModel):
     Year: int
-    SHAPE_Length: float
+    start_lat: float
+    start_lon: float
     dist_to_nearest_ev_m: float
     ev_within_500m: int
     avg_temp: float
@@ -28,11 +52,22 @@ class StationFeatures(BaseModel):
 def root():
     return {"message": "EV model API is running"}
 
-@app.post("/predict")
-def predict(features: StationFeatures):
-    data = pd.DataFrame([[getattr(features, f) for f in FEATURES]],
-                        columns=FEATURES)
+@app.post("/predict_from_coords")
+def predict_from_coords(req: CoordRequest):
+    shape_length = get_shape_length_from_coords(req.start_lat, req.start_lon)
+
+    data = pd.DataFrame([[
+        req.Year,
+        shape_length,
+        req.dist_to_nearest_ev_m,
+        req.ev_within_500m,
+        req.avg_temp,
+        req.total_prcp,
+    ]], columns=FEATURES)
 
     pred = model.predict(data)[0]
 
-    return {"prediction": float(pred)}
+    return {
+        "prediction": float(pred),
+        "used_SHAPE_Length": shape_length
+    }