adding the hf hub and dataloaders to gather the data from a private hf dataset

requirements.txt

@@ -2,4 +2,6 @@ altair
 pandas
 streamlit
 folium==0.16.0
-streamlit-folium==0.18.0
+streamlit-folium==0.18.0
+huggingface_hub>=0.21
+datasets

src/services/api.py

@@ -1,11 +1,25 @@
-import requests
+import os
+from typing import List, Dict, Any
+from datasets import load_dataset
 
-BACKEND_URL = "http://backend:8000"  # Use "http://localhost:8000" for local dev
+def get_visits_by_date(date: str) -> List[Dict[str, Any]]:
+    hf_token = os.getenv("HF_TOKEN")
+    if not hf_token:
+        return []
 
-def get_visits_by_date(date):
     try:
-        resp = requests.get(f"{BACKEND_URL}/api/visits", params={"end_date": date})
-        resp.raise_for_status()
-        return resp.json().get("data", [])
+        # Load the dataset from the specific branch
+        ds = load_dataset(
+            "uriels-locations/visits",
+            data_files="visitas_business_days_by_day_sample.json",
+            revision="refs/pr/2",
+            split="train",
+            use_auth_token=hf_token
+        )
     except Exception as e:
+        # Log error as needed
         return []
+
+    # Filter by date (assuming 'date' field exists)
+    filtered = [item for item in ds if item.get("date") == date]
+    return filtered