feat: add streamlit dashboard for paper trading and analytics

Author: VisionExpo

frontend/app.py

@@ -1,22 +1,22 @@
 import streamlit as st
+import requests
+from datetime import datetime
 
-st.set_page_config(
-    page_title="AI Stock Predictor - Home",
-    page_icon="🏠",
-    layout="wide"
-)
+API_BASE = st.secrets.get("API_BASE", "http://localhost:8000")
+
+st.set_page_config(page_title="ArthaQuant", layout="wide")
 
-st.title("🏠 Welcome to the AI Stock Predictor")
+st.title("📈 ArthaQuant — Paper Trading Dashboard")
 
-st.markdown("""
-This application is an end-to-end MLOps project for stock price prediction.
-It uses a containerized architecture with Docker, a FastAPI backend, and this Streamlit frontend.
+user_id = st.sidebar.number_input("User ID", min_value=1, value=1)
+
+page = st.sidebar.selectbox(
+    "Navigate",
+    ["Predict", "Paper Trade", "Portfolio", "Analytics", "Drift"],
+)
 
-**Features:**
-- **Predictions**: Get next-day price predictions for individual stocks using a Transformer model.
-- **Screener**: Run the model on multiple stocks to find potential opportunities.
-- **Drift Analysis**: Monitor for data drift between training and recent data.
-- **Backtesting**: (Coming Soon) Simulate trading strategies based on model predictions.
+def api_get(path):
+    return requests.get(f"{API_BASE}{path}").json()
 
-**Navigate to a feature using the sidebar on the left.**
-""")
+def api_post(path, payload):
+    return requests.post(f"{API_BASE}{path}", json=payload).json()

frontend/pages/__init__.py

@@ -0,0 +1,12 @@
+import streamlit as st
+from frontend.app import api_get
+
+st.header("📈 Analytics")
+
+data = api_get("/analytics")
+
+st.metric("Sharpe Ratio", data["sharpe_ratio"])
+st.metric("Max Drawdown", data["max_drawdown"])
+st.metric("Total Return", data["total_return"])
+
+st.line_chart(data["equity_curve"])

frontend/pages/analytics.py

@@ -0,0 +1,16 @@
+import streamlit as st
+import random
+from frontend.app import api_post
+
+st.header("🚨 Drift Detection")
+
+if st.button("Check Drift"):
+    reference = [random.random() for _ in range(100)]
+    current = [random.random() for _ in range(100)]
+
+    out = api_post("/drift", {
+        "reference": reference,
+        "current": current,
+    })
+
+    st.json(out)

frontend/pages/drift.py

@@ -0,0 +1,25 @@
+import streamlit as st
+from datetime import datetime
+from frontend.app import api_post
+
+st.header("💼 Paper Trade")
+
+symbol = st.text_input("Symbol", "AAPL")
+price = st.number_input("Current Price", value=100.0)
+
+p_up = st.slider("P(Up)", 0.0, 1.0, 0.75)
+expected_return = st.number_input("Expected Return", value=0.02)
+uncertainty = st.number_input("Uncertainty", value=0.1)
+
+if st.button("Execute Paper Trade"):
+    payload = {
+        "symbol": symbol,
+        "p_up": p_up,
+        "expected_return": expected_return,
+        "uncertainty": uncertainty,
+        "price": price,
+        "timestamp": datetime.utcnow().isoformat(),
+    }
+
+    out = api_post("/paper/trade", payload)
+    st.json(out)

frontend/pages/paper_trade.py

@@ -0,0 +1,15 @@
+import streamlit as st
+from frontend.app import api_get
+
+st.header("📊 Portfolio")
+
+data = api_get("/portfolio")
+
+st.subheader("Cash")
+st.write(data["cash"])
+
+st.subheader("Positions")
+st.json(data["positions"])
+
+st.subheader("Trades")
+st.json(data["trades"])

frontend/pages/portfolio.py

@@ -0,0 +1,23 @@
+import streamlit as st
+import numpy as np
+from frontend.app import api_post
+
+st.header("🔮 Predict")
+
+symbol = st.text_input("Symbol", "AAPL")
+
+if st.button("Run Prediction"):
+    market_seq = np.random.randn(60, 10).tolist()
+
+    payload = {
+        "symbol": symbol,
+        "market_sequence": market_seq,
+        "input_ids": [101] * 128,
+        "attention_mask": [1] * 128,
+    }
+
+    out = api_post("/predict", payload)
+
+    st.metric("P(Up)", f"{out['p_up']:.2f}")
+    st.metric("Expected Return", f"{out['expected_return']:.4f}")
+    st.metric("Uncertainty", f"{out['uncertainty']:.4f}")