import MetaTrader5 as mt5
import pandas as pd
import numpy as np
import plotly.graph_objects as go
from datetime import datetime, timedelta

# -------------------------
# 1. Initialize MT5
# -------------------------
if not mt5.initialize():
    print("MT5 initialization failed")
    mt5.shutdown()

# -------------------------
# 2. Fetch 3-minute data
# -------------------------
symbol = "XAUUSDc"
timeframe = mt5.TIMEFRAME_M3
n_bars = 500
utc_from = datetime.now() - timedelta(days=3)

rates = mt5.copy_rates_from(symbol, timeframe, utc_from, n_bars)
mt5.shutdown()

# -------------------------
# 3. Convert to DataFrame
# -------------------------
df = pd.DataFrame(rates)
df['time'] = pd.to_datetime(df['time'], unit='s')
df.set_index('time', inplace=True)

# -------------------------
# 4. Remove non-trading periods (weekends)
# -------------------------
df = df[df.index.dayofweek < 5]

# -------------------------
# 5. Feature engineering for VSA
# -------------------------
# basic bar measurements
df['range'] = df['high'] - df['low']
df['body'] = (df['close'] - df['open']).abs()
df['upper_wick'] = df['high'] - df[['open', 'close']].max(axis=1)
df['lower_wick'] = df[['open', 'close']].min(axis=1) - df['low']
df['direction'] = np.where(df['close'] > df['open'], 1, np.where(df['close'] < df['open'], -1, 0))

# rolling stats
vol_ma_len = 20
df['vol_ma'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).mean()
df['vol_std'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).std().fillna(0)
df['body_ma'] = df['body'].rolling(vol_ma_len, min_periods=1).mean()
df['range_ma'] = df['range'].rolling(vol_ma_len, min_periods=1).mean()

# normalized measures
df['vol_z'] = (df['tick_volume'] - df['vol_ma']) / (df['vol_std'] + 1e-9)
df['body_pct'] = df['body'] / (df['range_ma'] + 1e-9)

# -------------------------
# 6. VSA signal logic
# -------------------------
# signals: 1 buy signal, -1 sell signal, 0 none
df['signal'] = 0
df['signal_reason'] = ""

# parameters (tweakable)
climax_multiplier = 3.0   # very high volume relative to MA
high_vol_multiplier = 1.5 # moderately high volume
low_vol_multiplier = 0.5  # low volume threshold
small_body_thresh = 0.35  # body percent threshold to call "narrow body"
large_body_thresh = 0.6   # large body relative to range

# helper for trend context: short-term slope of closes
df['close_slope'] = df['close'].diff().rolling(6, min_periods=1).mean()

for i in range(2, len(df)):
    vol = df['tick_volume'].iat[i]
    vol_ma = df['vol_ma'].iat[i]
    direction = df['direction'].iat[i]
    body_pct = df['body_pct'].iat[i]
    upper_wick = df['upper_wick'].iat[i]
    lower_wick = df['lower_wick'].iat[i]
    close = df['close'].iat[i]
    open_ = df['open'].iat[i]
    prev_close = df['close'].iat[i-1]
    slope = df['close_slope'].iat[i-1]  # slope before this bar -> context

    # Safety: skip if nan
    if np.isnan(vol_ma) or vol_ma <= 0:
        continue

    # climactic volume up / down (effort producing strong result)
    if vol > vol_ma * climax_multiplier:
        if direction == 1 and body_pct > large_body_thresh:
            # climactic up: supply may be entering (potential reversal)
            df['signal'].iat[i] = -1
            df['signal_reason'].iat[i] = "Climax Up (high vol & large up bar) -> possible distribution"
            continue
        if direction == -1 and body_pct > large_body_thresh:
            # climactic down: selling climax -> possible exhaustion -> buy on test
            df['signal'].iat[i] = 1
            df['signal_reason'].iat[i] = "Climax Down (high vol & large down bar) -> possible selling exhaustion"
            continue

    # high volume down bar after uptrend (effort vs result)
    # If short-term prior slope positive and we now see high vol down close: weakness -> sell
    if vol > vol_ma * high_vol_multiplier and direction == -1 and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "High vol down after uptrend -> weakness (sell)"
        continue

    # high volume up bar after downtrend: strength -> buy
    if vol > vol_ma * high_vol_multiplier and direction == 1 and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "High vol up after downtrend -> strength (buy)"
        continue

    # No Demand (classic VSA): narrow up bar on low volume after advance -> likely no follow-through -> sell
    if vol < vol_ma * low_vol_multiplier and direction == 1 and body_pct < small_body_thresh and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "No Demand (low vol narrow up bar after advance) -> probable weak continuation"
        continue

    # No Supply: narrow down bar on low volume after decline -> likely no additional selling -> buy
    if vol < vol_ma * low_vol_multiplier and direction == -1 and body_pct < small_body_thresh and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "No Supply (low vol narrow down bar after decline) -> probable test, buy"
        continue

    # Testing (low vol down bar after selling climax) -> look for buy
    # If previous bar was a climactic down (very high vol down) and current bar is low vol with small down body => test => buy
    if i >= 2:
        prev_vol = df['tick_volume'].iat[i-1]
        prev_dir = df['direction'].iat[i-1]
        if prev_vol > df['vol_ma'].iat[i-1] * climax_multiplier and prev_dir == -1:
            if vol < vol_ma * low_vol_multiplier and direction != -1:
                df['signal'].iat[i] = 1
                df['signal_reason'].iat[i] = "Test after selling climax -> buy"
                continue

    # No signal default: continue loop

# -------------------------
# 7. Markers for plotting
# -------------------------
buy_mask = df['signal'] == 1
sell_mask = df['signal'] == -1

# price markers (above/below bars)
price_buy_y = df['low'] - (df['range'] * 0.15)
price_sell_y = df['high'] + (df['range'] * 0.15)

# volume marker heights
vol_max = df['tick_volume'].max()
vol_buy_y = df['tick_volume'][buy_mask]
vol_sell_y = df['tick_volume'][sell_mask]

# -------------------------
# 8. Plot 3-pane chart with signals
# -------------------------
fig = go.Figure()

# Candlestick
fig.add_trace(go.Candlestick(
    x=df.index,
    open=df['open'],
    high=df['high'],
    low=df['low'],
    close=df['close'],
    name=symbol,
    increasing_line_color='green',
    decreasing_line_color='red'
))

# Buy markers on price
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=price_buy_y[buy_mask],
    mode='markers',
    marker=dict(symbol='triangle-up', size=10),
    name='Buy Signal',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))

# Sell markers on price
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=price_sell_y[sell_mask],
    mode='markers',
    marker=dict(symbol='triangle-down', size=10),
    name='Sell Signal',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Tick-volume bars (separate y-axis)
fig.add_trace(go.Bar(
    x=df.index,
    y=df['tick_volume'],
    name='Tick Volume',
    marker_color='gray',
    yaxis='y2'
))

# Volume markers (overlay small markers on top of bars)
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=vol_buy_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Buy Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=vol_sell_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Sell Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Layout
fig.update_layout(
    template='plotly_white',
    title=f'{symbol} 3-Minute Candlestick with VSA Signals',
    xaxis=dict(title='Time', rangeslider=dict(visible=False)),
    yaxis=dict(title='Price', domain=[0.35, 1]),
    yaxis2=dict(title='Tick Volume', domain=[0.15, 0.33], showgrid=False, anchor="x"),
    legend=dict(orientation='h', y=1.02, x=0),
    height=900
)

fig.show()

# -------------------------
# 9. Quick signal summary
# -------------------------
recent_signals = df[df['signal'] != 0].copy().tail(20)
if not recent_signals.empty:
    print("Recent VSA signals (most recent first):")
    for idx, row in recent_signals[::-1].iterrows():
        ts = idx.strftime("%Y-%m-%d %H:%M")
        print(f"{ts} | Signal: {'BUY' if row['signal']==1 else 'SELL'} | Vol: {int(row['tick_volume'])} | Reason: {row['signal_reason']}")
else:
    print("No VSA signals found in the dataset.")

import MetaTrader5 as mt5
import pandas as pd
import numpy as np
import plotly.graph_objects as go
from datetime import datetime, timedelta

# -------------------------
# 1. Initialize MT5
# -------------------------
if not mt5.initialize():
    print("MT5 initialization failed")
    mt5.shutdown()

# -------------------------
# 2. Fetch 3-minute data
# -------------------------
symbol = "XAUUSDc"
timeframe = mt5.TIMEFRAME_M3
n_bars = 500
utc_from = datetime.now() - timedelta(days=3)

rates = mt5.copy_rates_from(symbol, timeframe, utc_from, n_bars)
mt5.shutdown()

# -------------------------
# 3. Convert to DataFrame
# -------------------------
df = pd.DataFrame(rates)
df['time'] = pd.to_datetime(df['time'], unit='s')
df.set_index('time', inplace=True)

# -------------------------
# 4. Remove non-trading periods (weekends)
# -------------------------
df = df[df.index.dayofweek < 5]

# -------------------------
# 5. Feature engineering for VSA
# -------------------------
# basic bar measurements
df['range'] = df['high'] - df['low']
df['body'] = (df['close'] - df['open']).abs()
df['upper_wick'] = df['high'] - df[['open', 'close']].max(axis=1)
df['lower_wick'] = df[['open', 'close']].min(axis=1) - df['low']
df['direction'] = np.where(df['close'] > df['open'], 1, np.where(df['close'] < df['open'], -1, 0))

# rolling stats
vol_ma_len = 20
df['vol_ma'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).mean()
df['vol_std'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).std().fillna(0)
df['body_ma'] = df['body'].rolling(vol_ma_len, min_periods=1).mean()
df['range_ma'] = df['range'].rolling(vol_ma_len, min_periods=1).mean()

# normalized measures
df['vol_z'] = (df['tick_volume'] - df['vol_ma']) / (df['vol_std'] + 1e-9)
df['body_pct'] = df['body'] / (df['range_ma'] + 1e-9)

# -------------------------
# 6. VSA signal logic
# -------------------------
# signals: 1 buy signal, -1 sell signal, 0 none
df['signal'] = 0
df['signal_reason'] = ""

# parameters (tweakable)
climax_multiplier = 3.0   # very high volume relative to MA
high_vol_multiplier = 1.5 # moderately high volume
low_vol_multiplier = 0.5  # low volume threshold
small_body_thresh = 0.35  # body percent threshold to call "narrow body"
large_body_thresh = 0.6   # large body relative to range

# helper for trend context: short-term slope of closes
df['close_slope'] = df['close'].diff().rolling(6, min_periods=1).mean()

for i in range(2, len(df)):
    vol = df['tick_volume'].iat[i]
    vol_ma = df['vol_ma'].iat[i]
    direction = df['direction'].iat[i]
    body_pct = df['body_pct'].iat[i]
    upper_wick = df['upper_wick'].iat[i]
    lower_wick = df['lower_wick'].iat[i]
    close = df['close'].iat[i]
    open_ = df['open'].iat[i]
    prev_close = df['close'].iat[i-1]
    slope = df['close_slope'].iat[i-1]  # slope before this bar -> context

    # Safety: skip if nan
    if np.isnan(vol_ma) or vol_ma <= 0:
        continue

    # climactic volume up / down (effort producing strong result)
    if vol > vol_ma * climax_multiplier:
        if direction == 1 and body_pct > large_body_thresh:
            # climactic up: supply may be entering (potential reversal)
            df['signal'].iat[i] = -1
            df['signal_reason'].iat[i] = "Climax Up (high vol & large up bar) -> possible distribution"
            continue
        if direction == -1 and body_pct > large_body_thresh:
            # climactic down: selling climax -> possible exhaustion -> buy on test
            df['signal'].iat[i] = 1
            df['signal_reason'].iat[i] = "Climax Down (high vol & large down bar) -> possible selling exhaustion"
            continue

    # high volume down bar after uptrend (effort vs result)
    # If short-term prior slope positive and we now see high vol down close: weakness -> sell
    if vol > vol_ma * high_vol_multiplier and direction == -1 and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "High vol down after uptrend -> weakness (sell)"
        continue

    # high volume up bar after downtrend: strength -> buy
    if vol > vol_ma * high_vol_multiplier and direction == 1 and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "High vol up after downtrend -> strength (buy)"
        continue

    # No Demand (classic VSA): narrow up bar on low volume after advance -> likely no follow-through -> sell
    if vol < vol_ma * low_vol_multiplier and direction == 1 and body_pct < small_body_thresh and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "No Demand (low vol narrow up bar after advance) -> probable weak continuation"
        continue

    # No Supply: narrow down bar on low volume after decline -> likely no additional selling -> buy
    if vol < vol_ma * low_vol_multiplier and direction == -1 and body_pct < small_body_thresh and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "No Supply (low vol narrow down bar after decline) -> probable test, buy"
        continue

    # Testing (low vol down bar after selling climax) -> look for buy
    # If previous bar was a climactic down (very high vol down) and current bar is low vol with small down body => test => buy
    if i >= 2:
        prev_vol = df['tick_volume'].iat[i-1]
        prev_dir = df['direction'].iat[i-1]
        if prev_vol > df['vol_ma'].iat[i-1] * climax_multiplier and prev_dir == -1:
            if vol < vol_ma * low_vol_multiplier and direction != -1:
                df['signal'].iat[i] = 1
                df['signal_reason'].iat[i] = "Test after selling climax -> buy"
                continue

    # No signal default: continue loop

# -------------------------
# 7. Markers for plotting
# -------------------------
buy_mask = df['signal'] == 1
sell_mask = df['signal'] == -1

# price markers (above/below bars)
price_buy_y = df['low'] - (df['range'] * 0.15)
price_sell_y = df['high'] + (df['range'] * 0.15)

# volume marker heights
vol_max = df['tick_volume'].max()
vol_buy_y = df['tick_volume'][buy_mask]
vol_sell_y = df['tick_volume'][sell_mask]

# -------------------------
# 8. Plot 3-pane chart with signals
# -------------------------
fig = go.Figure()

# Candlestick
fig.add_trace(go.Candlestick(
    x=df.index,
    open=df['open'],
    high=df['high'],
    low=df['low'],
    close=df['close'],
    name=symbol,
    increasing_line_color='green',
    decreasing_line_color='red'
))

# Buy markers on price
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=price_buy_y[buy_mask],
    mode='markers',
    marker=dict(symbol='triangle-up', size=10),
    name='Buy Signal',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))

# Sell markers on price
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=price_sell_y[sell_mask],
    mode='markers',
    marker=dict(symbol='triangle-down', size=10),
    name='Sell Signal',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Tick-volume bars (separate y-axis)
fig.add_trace(go.Bar(
    x=df.index,
    y=df['tick_volume'],
    name='Tick Volume',
    marker_color='gray',
    yaxis='y2'
))

# Volume markers (overlay small markers on top of bars)
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=vol_buy_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Buy Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=vol_sell_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Sell Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Layout
fig.update_layout(
    template='plotly_white',
    title=f'{symbol} 3-Minute Candlestick with VSA Signals',
    xaxis=dict(title='Time', rangeslider=dict(visible=False)),
    yaxis=dict(title='Price', domain=[0.35, 1]),
    yaxis2=dict(title='Tick Volume', domain=[0.15, 0.33], showgrid=False, anchor="x"),
    legend=dict(orientation='h', y=1.02, x=0),
    height=900
)

fig.show()

# -------------------------
# 9. Quick signal summary
# -------------------------
recent_signals = df[df['signal'] != 0].copy().tail(20)
if not recent_signals.empty:
    print("Recent VSA signals (most recent first):")
    for idx, row in recent_signals[::-1].iterrows():
        ts = idx.strftime("%Y-%m-%d %H:%M")
        print(f"{ts} | Signal: {'BUY' if row['signal']==1 else 'SELL'} | Vol: {int(row['tick_volume'])} | Reason: {row['signal_reason']}")
else:
    print("No VSA signals found in the dataset.")

import MetaTrader5 as mt5
import pandas as pd
import numpy as np
import plotly.graph_objects as go
from datetime import datetime, timedelta

# -------------------------
# 1. Initialize MT5
# -------------------------
if not mt5.initialize():
    print("MT5 initialization failed")
    mt5.shutdown()

# -------------------------
# 2. Fetch 3-minute data
# -------------------------
symbol = "XAUUSDc"
timeframe = mt5.TIMEFRAME_M3
n_bars = 500
utc_from = datetime.now() - timedelta(days=3)

rates = mt5.copy_rates_from(symbol, timeframe, utc_from, n_bars)
mt5.shutdown()

# -------------------------
# 3. Convert to DataFrame
# -------------------------
df = pd.DataFrame(rates)
df['time'] = pd.to_datetime(df['time'], unit='s')
df.set_index('time', inplace=True)

# -------------------------
# 4. Remove non-trading periods (weekends)
# -------------------------
df = df[df.index.dayofweek < 5]

# -------------------------
# 5. Feature engineering for VSA
# -------------------------
# basic bar measurements
df['range'] = df['high'] - df['low']
df['body'] = (df['close'] - df['open']).abs()
df['upper_wick'] = df['high'] - df[['open', 'close']].max(axis=1)
df['lower_wick'] = df[['open', 'close']].min(axis=1) - df['low']
df['direction'] = np.where(df['close'] > df['open'], 1, np.where(df['close'] < df['open'], -1, 0))

# rolling stats
vol_ma_len = 20
df['vol_ma'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).mean()
df['vol_std'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).std().fillna(0)
df['body_ma'] = df['body'].rolling(vol_ma_len, min_periods=1).mean()
df['range_ma'] = df['range'].rolling(vol_ma_len, min_periods=1).mean()

# normalized measures
df['vol_z'] = (df['tick_volume'] - df['vol_ma']) / (df['vol_std'] + 1e-9)
df['body_pct'] = df['body'] / (df['range_ma'] + 1e-9)

# -------------------------
# 6. VSA signal logic
# -------------------------
# signals: 1 buy signal, -1 sell signal, 0 none
df['signal'] = 0
df['signal_reason'] = ""

# parameters (tweakable)
climax_multiplier = 3.0   # very high volume relative to MA
high_vol_multiplier = 1.5 # moderately high volume
low_vol_multiplier = 0.5  # low volume threshold
small_body_thresh = 0.35  # body percent threshold to call "narrow body"
large_body_thresh = 0.6   # large body relative to range

# helper for trend context: short-term slope of closes
df['close_slope'] = df['close'].diff().rolling(6, min_periods=1).mean()

for i in range(2, len(df)):
    vol = df['tick_volume'].iat[i]
    vol_ma = df['vol_ma'].iat[i]
    direction = df['direction'].iat[i]
    body_pct = df['body_pct'].iat[i]
    upper_wick = df['upper_wick'].iat[i]
    lower_wick = df['lower_wick'].iat[i]
    close = df['close'].iat[i]
    open_ = df['open'].iat[i]
    prev_close = df['close'].iat[i-1]
    slope = df['close_slope'].iat[i-1]  # slope before this bar -> context

    # Safety: skip if nan
    if np.isnan(vol_ma) or vol_ma <= 0:
        continue

    # climactic volume up / down (effort producing strong result)
    if vol > vol_ma * climax_multiplier:
        if direction == 1 and body_pct > large_body_thresh:
            # climactic up: supply may be entering (potential reversal)
            df['signal'].iat[i] = -1
            df['signal_reason'].iat[i] = "Climax Up (high vol & large up bar) -> possible distribution"
            continue
        if direction == -1 and body_pct > large_body_thresh:
            # climactic down: selling climax -> possible exhaustion -> buy on test
            df['signal'].iat[i] = 1
            df['signal_reason'].iat[i] = "Climax Down (high vol & large down bar) -> possible selling exhaustion"
            continue

    # high volume down bar after uptrend (effort vs result)
    # If short-term prior slope positive and we now see high vol down close: weakness -> sell
    if vol > vol_ma * high_vol_multiplier and direction == -1 and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "High vol down after uptrend -> weakness (sell)"
        continue

    # high volume up bar after downtrend: strength -> buy
    if vol > vol_ma * high_vol_multiplier and direction == 1 and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "High vol up after downtrend -> strength (buy)"
        continue

    # No Demand (classic VSA): narrow up bar on low volume after advance -> likely no follow-through -> sell
    if vol < vol_ma * low_vol_multiplier and direction == 1 and body_pct < small_body_thresh and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "No Demand (low vol narrow up bar after advance) -> probable weak continuation"
        continue

    # No Supply: narrow down bar on low volume after decline -> likely no additional selling -> buy
    if vol < vol_ma * low_vol_multiplier and direction == -1 and body_pct < small_body_thresh and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "No Supply (low vol narrow down bar after decline) -> probable test, buy"
        continue

    # Testing (low vol down bar after selling climax) -> look for buy
    # If previous bar was a climactic down (very high vol down) and current bar is low vol with small down body => test => buy
    if i >= 2:
        prev_vol = df['tick_volume'].iat[i-1]
        prev_dir = df['direction'].iat[i-1]
        if prev_vol > df['vol_ma'].iat[i-1] * climax_multiplier and prev_dir == -1:
            if vol < vol_ma * low_vol_multiplier and direction != -1:
                df['signal'].iat[i] = 1
                df['signal_reason'].iat[i] = "Test after selling climax -> buy"
                continue

    # No signal default: continue loop

# -------------------------
# 7. Markers for plotting
# -------------------------
buy_mask = df['signal'] == 1
sell_mask = df['signal'] == -1

# price markers (above/below bars)
price_buy_y = df['low'] - (df['range'] * 0.15)
price_sell_y = df['high'] + (df['range'] * 0.15)

# volume marker heights
vol_max = df['tick_volume'].max()
vol_buy_y = df['tick_volume'][buy_mask]
vol_sell_y = df['tick_volume'][sell_mask]

# -------------------------
# 8. Plot 3-pane chart with signals
# -------------------------
fig = go.Figure()

# Candlestick
fig.add_trace(go.Candlestick(
    x=df.index,
    open=df['open'],
    high=df['high'],
    low=df['low'],
    close=df['close'],
    name=symbol,
    increasing_line_color='green',
    decreasing_line_color='red'
))

# Buy markers on price
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=price_buy_y[buy_mask],
    mode='markers',
    marker=dict(symbol='triangle-up', size=10),
    name='Buy Signal',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))

# Sell markers on price
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=price_sell_y[sell_mask],
    mode='markers',
    marker=dict(symbol='triangle-down', size=10),
    name='Sell Signal',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Tick-volume bars (separate y-axis)
fig.add_trace(go.Bar(
    x=df.index,
    y=df['tick_volume'],
    name='Tick Volume',
    marker_color='gray',
    yaxis='y2'
))

# Volume markers (overlay small markers on top of bars)
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=vol_buy_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Buy Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=vol_sell_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Sell Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Layout
fig.update_layout(
    template='plotly_white',
    title=f'{symbol} 3-Minute Candlestick with VSA Signals',
    xaxis=dict(title='Time', rangeslider=dict(visible=False)),
    yaxis=dict(title='Price', domain=[0.35, 1]),
    yaxis2=dict(title='Tick Volume', domain=[0.15, 0.33], showgrid=False, anchor="x"),
    legend=dict(orientation='h', y=1.02, x=0),
    height=900
)

fig.show()

# -------------------------
# 9. Quick signal summary
# -------------------------
recent_signals = df[df['signal'] != 0].copy().tail(20)
if not recent_signals.empty:
    print("Recent VSA signals (most recent first):")
    for idx, row in recent_signals[::-1].iterrows():
        ts = idx.strftime("%Y-%m-%d %H:%M")
        print(f"{ts} | Signal: {'BUY' if row['signal']==1 else 'SELL'} | Vol: {int(row['tick_volume'])} | Reason: {row['signal_reason']}")
else:
    print("No VSA signals found in the dataset.")

import MetaTrader5 as mt5
import pandas as pd
import numpy as np
import plotly.graph_objects as go
from datetime import datetime, timedelta

# -------------------------
# 1. Initialize MT5
# -------------------------
if not mt5.initialize():
    print("MT5 initialization failed")
    mt5.shutdown()

# -------------------------
# 2. Fetch 3-minute data
# -------------------------
symbol = "XAUUSDc"
timeframe = mt5.TIMEFRAME_M3
n_bars = 500
utc_from = datetime.now() - timedelta(days=3)

rates = mt5.copy_rates_from(symbol, timeframe, utc_from, n_bars)
mt5.shutdown()

# -------------------------
# 3. Convert to DataFrame
# -------------------------
df = pd.DataFrame(rates)
df['time'] = pd.to_datetime(df['time'], unit='s')
df.set_index('time', inplace=True)

# -------------------------
# 4. Remove non-trading periods (weekends)
# -------------------------
df = df[df.index.dayofweek < 5]

# -------------------------
# 5. Feature engineering for VSA
# -------------------------
# basic bar measurements
df['range'] = df['high'] - df['low']
df['body'] = (df['close'] - df['open']).abs()
df['upper_wick'] = df['high'] - df[['open', 'close']].max(axis=1)
df['lower_wick'] = df[['open', 'close']].min(axis=1) - df['low']
df['direction'] = np.where(df['close'] > df['open'], 1, np.where(df['close'] < df['open'], -1, 0))

# rolling stats
vol_ma_len = 20
df['vol_ma'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).mean()
df['vol_std'] = df['tick_volume'].rolling(vol_ma_len, min_periods=1).std().fillna(0)
df['body_ma'] = df['body'].rolling(vol_ma_len, min_periods=1).mean()
df['range_ma'] = df['range'].rolling(vol_ma_len, min_periods=1).mean()

# normalized measures
df['vol_z'] = (df['tick_volume'] - df['vol_ma']) / (df['vol_std'] + 1e-9)
df['body_pct'] = df['body'] / (df['range_ma'] + 1e-9)

# -------------------------
# 6. VSA signal logic
# -------------------------
# signals: 1 buy signal, -1 sell signal, 0 none
df['signal'] = 0
df['signal_reason'] = ""

# parameters (tweakable)
climax_multiplier = 3.0   # very high volume relative to MA
high_vol_multiplier = 1.5 # moderately high volume
low_vol_multiplier = 0.5  # low volume threshold
small_body_thresh = 0.35  # body percent threshold to call "narrow body"
large_body_thresh = 0.6   # large body relative to range

# helper for trend context: short-term slope of closes
df['close_slope'] = df['close'].diff().rolling(6, min_periods=1).mean()

for i in range(2, len(df)):
    vol = df['tick_volume'].iat[i]
    vol_ma = df['vol_ma'].iat[i]
    direction = df['direction'].iat[i]
    body_pct = df['body_pct'].iat[i]
    upper_wick = df['upper_wick'].iat[i]
    lower_wick = df['lower_wick'].iat[i]
    close = df['close'].iat[i]
    open_ = df['open'].iat[i]
    prev_close = df['close'].iat[i-1]
    slope = df['close_slope'].iat[i-1]  # slope before this bar -> context

    # Safety: skip if nan
    if np.isnan(vol_ma) or vol_ma <= 0:
        continue

    # climactic volume up / down (effort producing strong result)
    if vol > vol_ma * climax_multiplier:
        if direction == 1 and body_pct > large_body_thresh:
            # climactic up: supply may be entering (potential reversal)
            df['signal'].iat[i] = -1
            df['signal_reason'].iat[i] = "Climax Up (high vol & large up bar) -> possible distribution"
            continue
        if direction == -1 and body_pct > large_body_thresh:
            # climactic down: selling climax -> possible exhaustion -> buy on test
            df['signal'].iat[i] = 1
            df['signal_reason'].iat[i] = "Climax Down (high vol & large down bar) -> possible selling exhaustion"
            continue

    # high volume down bar after uptrend (effort vs result)
    # If short-term prior slope positive and we now see high vol down close: weakness -> sell
    if vol > vol_ma * high_vol_multiplier and direction == -1 and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "High vol down after uptrend -> weakness (sell)"
        continue

    # high volume up bar after downtrend: strength -> buy
    if vol > vol_ma * high_vol_multiplier and direction == 1 and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "High vol up after downtrend -> strength (buy)"
        continue

    # No Demand (classic VSA): narrow up bar on low volume after advance -> likely no follow-through -> sell
    if vol < vol_ma * low_vol_multiplier and direction == 1 and body_pct < small_body_thresh and slope > 0:
        df['signal'].iat[i] = -1
        df['signal_reason'].iat[i] = "No Demand (low vol narrow up bar after advance) -> probable weak continuation"
        continue

    # No Supply: narrow down bar on low volume after decline -> likely no additional selling -> buy
    if vol < vol_ma * low_vol_multiplier and direction == -1 and body_pct < small_body_thresh and slope < 0:
        df['signal'].iat[i] = 1
        df['signal_reason'].iat[i] = "No Supply (low vol narrow down bar after decline) -> probable test, buy"
        continue

    # Testing (low vol down bar after selling climax) -> look for buy
    # If previous bar was a climactic down (very high vol down) and current bar is low vol with small down body => test => buy
    if i >= 2:
        prev_vol = df['tick_volume'].iat[i-1]
        prev_dir = df['direction'].iat[i-1]
        if prev_vol > df['vol_ma'].iat[i-1] * climax_multiplier and prev_dir == -1:
            if vol < vol_ma * low_vol_multiplier and direction != -1:
                df['signal'].iat[i] = 1
                df['signal_reason'].iat[i] = "Test after selling climax -> buy"
                continue

    # No signal default: continue loop

# -------------------------
# 7. Markers for plotting
# -------------------------
buy_mask = df['signal'] == 1
sell_mask = df['signal'] == -1

# price markers (above/below bars)
price_buy_y = df['low'] - (df['range'] * 0.15)
price_sell_y = df['high'] + (df['range'] * 0.15)

# volume marker heights
vol_max = df['tick_volume'].max()
vol_buy_y = df['tick_volume'][buy_mask]
vol_sell_y = df['tick_volume'][sell_mask]

# -------------------------
# 8. Plot 3-pane chart with signals
# -------------------------
fig = go.Figure()

# Candlestick
fig.add_trace(go.Candlestick(
    x=df.index,
    open=df['open'],
    high=df['high'],
    low=df['low'],
    close=df['close'],
    name=symbol,
    increasing_line_color='green',
    decreasing_line_color='red'
))

# Buy markers on price
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=price_buy_y[buy_mask],
    mode='markers',
    marker=dict(symbol='triangle-up', size=10),
    name='Buy Signal',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))

# Sell markers on price
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=price_sell_y[sell_mask],
    mode='markers',
    marker=dict(symbol='triangle-down', size=10),
    name='Sell Signal',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Tick-volume bars (separate y-axis)
fig.add_trace(go.Bar(
    x=df.index,
    y=df['tick_volume'],
    name='Tick Volume',
    marker_color='gray',
    yaxis='y2'
))

# Volume markers (overlay small markers on top of bars)
fig.add_trace(go.Scatter(
    x=df.index[buy_mask],
    y=vol_buy_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Buy Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][buy_mask],
    hoverinfo='text'
))
fig.add_trace(go.Scatter(
    x=df.index[sell_mask],
    y=vol_sell_y,
    mode='markers',
    marker=dict(symbol='circle', size=8),
    name='Volume Sell Marker',
    yaxis='y2',
    hovertext=df['signal_reason'][sell_mask],
    hoverinfo='text'
))

# Layout
fig.update_layout(
    template='plotly_white',
    title=f'{symbol} 3-Minute Candlestick with VSA Signals',
    xaxis=dict(title='Time', rangeslider=dict(visible=False)),
    yaxis=dict(title='Price', domain=[0.35, 1]),
    yaxis2=dict(title='Tick Volume', domain=[0.15, 0.33], showgrid=False, anchor="x"),
    legend=dict(orientation='h', y=1.02, x=0),
    height=900
)

fig.show()

# -------------------------
# 9. Quick signal summary
# -------------------------
recent_signals = df[df['signal'] != 0].copy().tail(20)
if not recent_signals.empty:
    print("Recent VSA signals (most recent first):")
    for idx, row in recent_signals[::-1].iterrows():
        ts = idx.strftime("%Y-%m-%d %H:%M")
        print(f"{ts} | Signal: {'BUY' if row['signal']==1 else 'SELL'} | Vol: {int(row['tick_volume'])} | Reason: {row['signal_reason']}")
else:
    print("No VSA signals found in the dataset.")