Create app_min.py

app_min.py

@@ -0,0 +1,41 @@
+
+# app_min.py
+import io, json, pandas as pd, numpy as np, matplotlib.pyplot as plt
+from PIL import Image
+import gradio as gr
+REQUIRED = {"date","revenue","cogs","opex"}
+
+def load_csv(file):
+    try: df = pd.read_csv(file)
+    except Exception as e: raise gr.Error(json.dumps({"error":"Failed to read CSV","details":str(e)}, indent=2))
+    df.columns = [c.strip().lower() for c in df.columns]
+    miss = REQUIRED - set(df.columns)
+    if miss: raise gr.Error(json.dumps({"error":"CSV schema invalid","details":f"Missing columns: {sorted(list(miss))}"} , indent=2))
+    try:
+        df["date"] = pd.to_datetime(df["date"]).dt.to_period("M").dt.to_timestamp()
+        for col in ["revenue","cogs","opex"]: df[col] = pd.to_numeric(df[col])
+    except Exception as e: raise gr.Error(json.dumps({"error":"Parsing failed","details":str(e)}, indent=2))
+    return df.sort_values("date").reset_index(drop=True)
+
+def plot(df):
+    fig, ax = plt.subplots(1,1, figsize=(8,4))
+    ax.plot(df["date"], df["revenue"], label="Revenue")
+    ax.plot(df["date"], df["cogs"], label="COGS")
+    ax.plot(df["date"], df["opex"], label="Opex")
+    ax.legend(); ax.set_title("P&L (minimal)")
+    buf = io.BytesIO(); plt.tight_layout(); fig.savefig(buf, format="png"); plt.close(fig)
+    return Image.open(buf)
+
+def run(file):
+    df = load_csv(file); img = plot(df)
+    return img, ("echo.csv", df.to_csv(index=False).encode()), json.dumps({"rows":len(df)}, indent=2)
+
+with gr.Blocks(title="Minimal Forecast Agent Test") as demo:
+    f = gr.File(file_types=[".csv"], label="Upload CSV")
+    btn = gr.Button("Run")
+    img = gr.Image(type="pil", label="Plot")
+    dl  = gr.File(label="Echo CSV")
+    meta= gr.JSON(label="Diagnostics")
+    btn.click(lambda x: run(x.name if x else None), inputs=f, outputs=[img, dl, meta])
+
+if __name__ == '__main__': demo.launch()