Update app.py

app.py

@@ -1,136 +1,115 @@
-import streamlit as st
-import numpy as np
-import pandas as pd
-import matplotlib.pyplot as plt
-
-st.set_page_config(page_title="AVOLD", layout="centered")
-
-# ------------------------
-# SESSION
-# ------------------------
-if "history" not in st.session_state:
-    st.session_state.history = []
-
-if "error_log" not in st.session_state:
-    st.session_state.error_log = []
-
-# ------------------------
-# TITLE
-# ------------------------
-st.markdown("<h1 style='text-align:center;'>AVOLD</h1>", unsafe_allow_html=True)
-
-# ------------------------
-# INPUT + RESET
-# ------------------------
-col1, col2 = st.columns([3,1])
-
-with col1:
-    val = st.number_input("", min_value=1.0, step=0.01, placeholder="1.45")
-
-with col2:
-    if st.button("RESET"):
-        st.session_state.history = []
-        st.session_state.error_log = []
-
-if st.button("ADD"):
-    st.session_state.history.append(round(val,2))
-
-data = st.session_state.history[-50:]
-
-if len(data) < 5:
-    st.stop()
-
-arr = np.array(data)
-
-# ------------------------
-# CORE METRICS
-# ------------------------
-mean = np.mean(arr)
-std = np.std(arr)
-
-low_ratio = np.sum(arr < 1.5)/len(arr)
-high_ratio = np.sum(arr > 3)/len(arr)
-
-chaos = (std * high_ratio) / (low_ratio + 0.1)
-
-# ------------------------
-# EXPECTED (ONLINE LEARNING EWMA)
-# ------------------------
-alpha = 0.3
-ewma = arr[0]
-for x in arr[1:]:
-    ewma = alpha*x + (1-alpha)*ewma
-
-median = np.median(arr)
-
-expected = (ewma*0.6) + (median*0.4)
-
-# range
-lower = max(1.01, expected - std*0.5)
-upper = expected + std*0.5
-
-# ------------------------
-# PROBABILITY ENGINE
-# ------------------------
-prob_2 = np.sum(arr>=2)/len(arr)
-prob_3 = np.sum(arr>=3)/len(arr)
-
-# ------------------------
-# ONLINE LEARNING ERROR UPDATE
-# ------------------------
-if len(arr) > 1:
-    last_expected = expected
-    last_actual = arr[-1]
-    error = abs(last_expected - last_actual)
-    st.session_state.error_log.append(error)
-
-avg_error = np.mean(st.session_state.error_log) if len(st.session_state.error_log)>0 else 0
-
-# adaptive confidence
-stability = max(0,1-std/5)
-confidence = max(0, min(1, (1-avg_error/3 + stability)/2 ))
-
-# ------------------------
-# DECISION
-# ------------------------
-if chaos > 1.2:
-    decision = "SKIP"
-elif expected > 2 and confidence > 0.6:
-    decision = "ENTER+"
-elif expected > 1.4 and confidence > 0.5:
-    decision = "ENTER"
-else:
-    decision = "WAIT"
-
-# ------------------------
-# OUTPUT BOX (ONE LINE)
-# ------------------------
-c1, c2, c3 = st.columns(3)
-
-c1.metric("Expected", round(expected,2))
-c2.metric("Analysis", f"{round(prob_2,2)} | {round(prob_3,2)}")
-c3.metric("Decision", decision)
-
-# ------------------------
-# CHART 1: TREND
-# ------------------------
-fig1, ax1 = plt.subplots()
-ax1.plot(arr, linewidth=2)
-ax1.set_title("Trend")
-ax1.set_ylabel("Multiplier")
-st.pyplot(fig1)
-
-# ------------------------
-# CHART 2: ERROR (ONLINE LEARNING)
-# ------------------------
-if len(st.session_state.error_log) > 3:
-    fig2, ax2 = plt.subplots()
-    ax2.plot(st.session_state.error_log[-50:], linewidth=2)
-    ax2.set_title("Learning Error")
-    st.pyplot(fig2)
-
-# ------------------------
-# BOTTOM DATA
-# ------------------------
-st.markdown("---")
-st.write(data)
+# app.py
+import gradio as gr
+import random
+from statistics import mean
+
+# ---------- Helper functions ----------
+def compute_outputs(rounds):
+    """Calculate expected multiplier, analysis, and decision based on recent rounds."""
+    if not rounds:
+        return "0.0", "No data", "N/A"
+
+    # Use last 10 rounds (or fewer if not enough data)
+    recent = rounds[-10:] if len(rounds) >= 10 else rounds
+    avg = mean(recent)
+
+    # Expected multiplier = average ± small random fluctuation (simulate unpredictability)
+    expected = round(avg * random.uniform(0.95, 1.05), 2)
+
+    # Simple trend analysis
+    if len(recent) >= 2:
+        trend = recent[-1] - recent[-2]
+        if trend > 0.1:
+            analysis = "Increasing"
+        elif trend < -0.1:
+            analysis = "Decreasing"
+        else:
+            analysis = "Stable"
+    else:
+        analysis = "Insufficient data"
+
+    # Decision based on expected multiplier
+    if expected > 2.5:
+        decision = "High"
+    elif expected > 1.5:
+        decision = "Medium"
+    else:
+        decision = "Low"
+
+    return str(expected), analysis, decision
+
+def update_ui(rounds):
+    """Prepare the dataframe and the three output values from the current rounds list."""
+    # Build table: [Round number, Multiplier]
+    table = [[i+1, val] for i, val in enumerate(rounds)]
+    exp, ana, dec = compute_outputs(rounds)
+    return table, exp, ana, dec
+
+def add_round(state, new_val):
+    """Append a new multiplier, keep only last 50, and refresh UI."""
+    if new_val <= 0:
+        # ignore invalid inputs (but you could show a warning – here we just return unchanged)
+        return state + update_ui(state)[:3]   # need to return state and outputs
+    rounds = state.copy()
+    rounds.append(new_val)
+    if len(rounds) > 50:
+        rounds = rounds[-50:]
+    table, exp, ana, dec = update_ui(rounds)
+    return rounds, table, exp, ana, dec
+
+def reset_rounds():
+    """Reset to a fresh set of 20 random multipliers."""
+    initial = [round(random.uniform(1.0, 3.0), 2) for _ in range(20)]
+    table, exp, ana, dec = update_ui(initial)
+    return initial, table, exp, ana, dec
+
+# ---------- Build the Gradio interface ----------
+with gr.Blocks(theme='dark', css="footer {visibility: hidden}") as demo:
+    gr.Markdown("# AVOLD")
+
+    # State: stores the list of multipliers
+    rounds_state = gr.State([round(random.uniform(1.0, 3.0), 2) for _ in range(20)])
+
+    # Input row
+    with gr.Row():
+        new_multiplier = gr.Number(label="New Multiplier", value=1.0, step=0.1)
+        add_btn = gr.Button("Add Round")
+
+    # Reset button
+    reset_btn = gr.Button("Reset Data")
+
+    # Three outputs in one row
+    with gr.Row():
+        expected_out = gr.Textbox(label="Expected Multiplier", interactive=False)
+        analysis_out = gr.Textbox(label="Analysis", interactive=False)
+        decision_out = gr.Textbox(label="Decision", interactive=False)
+
+    # Table at the bottom
+    rounds_table = gr.Dataframe(
+        label="Last 50 Rounds",
+        headers=["Round", "Multiplier"],
+        row_count=10,          # show 10 rows at a time (scrollable)
+    )
+
+    # Wire up the buttons
+    add_btn.click(
+        add_round,
+        inputs=[rounds_state, new_multiplier],
+        outputs=[rounds_state, rounds_table, expected_out, analysis_out, decision_out]
+    )
+
+    reset_btn.click(
+        reset_rounds,
+        outputs=[rounds_state, rounds_table, expected_out, analysis_out, decision_out]
+    )
+
+    # Initial load
+    demo.load(
+        lambda s: update_ui(s),
+        inputs=[rounds_state],
+        outputs=[rounds_table, expected_out, analysis_out, decision_out]
+    )
+
+# Launch the app
+if __name__ == "__main__":
+    demo.launch()