Update app.py

app.py

@@ -1,4 +1,3 @@
-import io
 import json
 import gradio as gr
 import pandas as pd
@@ -21,14 +20,16 @@ You can **generate a sample** dataset or **upload a CSV** with columns:
 - **Time windows** are *soft* (violations are reported in metrics).
 - Distances are Euclidean on the X-Y plane for clarity.
 
-Tip: Start with the generator, then switch to your CSV.
+💡 Tip: Start with the generator, then switch to your CSV.
 """
 
 FOOTER = "Made with ❤️ using Gradio. No native dependencies; runs quickly on Spaces."
 
-def run_generator(
-    n_clients, n_vehicles, capacity, spread, demand_min, demand_max, seed
-):
+
+# -----------------------------
+# Functions
+# -----------------------------
+def run_generator(n_clients, n_vehicles, capacity, spread, demand_min, demand_max, seed):
     df = generate_random_instance(
         n_clients=n_clients,
         n_vehicles=n_vehicles,
@@ -39,23 +40,18 @@ def run_generator(
         seed=seed,
     )
     depot = (0.0, 0.0)
-    sol = solve_vrp(
-        df, depot=depot, n_vehicles=n_vehicles, capacity=capacity, speed=1.0
-    )
-    fig = plot_solution(df, sol, depot=depot)
-    buf = io.BytesIO()
-    fig.savefig(buf, format="png", bbox_inches="tight", dpi=160)
-    buf.seek(0)
+    sol = solve_vrp(df, depot=depot, n_vehicles=n_vehicles, capacity=capacity, speed=1.0)
+
+    # ✅ plot_solution now returns a PIL image
+    img = plot_solution(df, sol, depot=depot)
 
-    # Route table
     route_table = sol["assignments_table"]
     metrics = json.dumps(sol["metrics"], indent=2)
 
-    return buf, route_table, metrics, df
+    return img, route_table, metrics, df
 
-def run_csv(
-    file, n_vehicles, capacity
-):
+
+def run_csv(file, n_vehicles, capacity):
     if file is None:
         raise gr.Error("Please upload a CSV first.")
     try:
@@ -64,22 +60,22 @@ def run_csv(
         raise gr.Error(f"CSV parsing error: {e}")
 
     depot = (0.0, 0.0)
-    sol = solve_vrp(
-        df, depot=depot, n_vehicles=n_vehicles, capacity=capacity, speed=1.0
-    )
-    fig = plot_solution(df, sol, depot=depot)
-    buf = io.BytesIO()
-    fig.savefig(buf, format="png", bbox_inches="tight", dpi=160)
-    buf.seek(0)
+    sol = solve_vrp(df, depot=depot, n_vehicles=n_vehicles, capacity=capacity, speed=1.0)
+
+    img = plot_solution(df, sol, depot=depot)
 
     route_table = sol["assignments_table"]
     metrics = json.dumps(sol["metrics"], indent=2)
-    return buf, route_table, metrics
+    return img, route_table, metrics
+
 
 def download_template():
-    df = make_template_dataframe()
-    return df
+    return make_template_dataframe()
 
+
+# -----------------------------
+# UI Layout
+# -----------------------------
 with gr.Blocks(title=TITLE) as demo:
     gr.Markdown(f"# {TITLE}")
     gr.Markdown(DESC)
@@ -96,7 +92,8 @@ with gr.Blocks(title=TITLE) as demo:
                 seed = gr.Slider(0, 9999, value=42, step=1, label="Random seed")
                 run_btn = gr.Button("🚗 Generate & Optimize", variant="primary")
             with gr.Column():
-                img = gr.Image(label="Route Visualization", interactive=False)
+                img = gr.Image(type="pil", label="Route Visualization", interactive=False)
+
         with gr.Row():
             route_df = gr.Dataframe(label="Route assignments (per stop)", wrap=True)
             metrics = gr.Code(label="Metrics (JSON)")
@@ -118,7 +115,7 @@ with gr.Blocks(title=TITLE) as demo:
                 capacity2 = gr.Slider(1, 200, value=15, step=1, label="Vehicle capacity (sum of demand)")
                 run_btn2 = gr.Button("📈 Optimize uploaded data", variant="primary")
             with gr.Column():
-                img2 = gr.Image(label="Route Visualization", interactive=False)
+                img2 = gr.Image(type="pil", label="Route Visualization", interactive=False)
 
         with gr.Row():
             route_df2 = gr.Dataframe(label="Route assignments (per stop)")