Flow-K: K-line Trend, Slope & Signal Analysis Tool

A focused technical analysis tool that detects trend lines, slope dynamics, MA curve patterns, and candlestick reversal patterns from real-time stock data, then visualizes everything with professional charts.

What It Does

• Fetches real-time OHLCV data via yfinance or local S&P500 CSV
• Computes MA5/MA10/MA20/MA60 and their slopes (linear regression angle)
• Detects trend lines (diagonal support/resistance) and price breakouts/bounces
• Identifies slope changes: acceleration, deceleration, bullish/bearish divergence
• Analyzes MA curves: golden/death cross, convergence/divergence, bullish/bearish alignment, Bollinger squeeze
• Detects classic candlestick patterns (enhanced with confirmation): 乌云盖顶, 刺透, 晨星, 黄昏之星, 锤头, 射击之星, 吞没, 白三兵, 三只乌鸦, etc.
• Generates a composite signal score (看多/看空/中性) based on all detected factors
• Generates professional candlestick charts with trend lines, MA curves, signal markers

Quick Start

pip install -r requirements.txt

# CLI
python src/analyzer.py NVDA

# Backtest a specific strategy
python src/analyzer.py NVDA --strategy full_monty

# Web dashboard
python src/api.py
# Open http://localhost:5000

CLI Usage

python src/analyzer.py NVDA              # Default: 1 year daily
python src/analyzer.py NVDA --period 6mo  # 6 months
python src/analyzer.py 0700.HK            # HK stock
python src/analyzer.py AAPL --mas 5,10,20,60

Web Dashboard

python src/api.py

Interactive dashboard with:

• Stock ticker input + period selector
• Overall trend card (上升 / 下降 / 横盘)
• MA alignment card (多头排列 / 空头排列 / 交织)
• Signal score card with bullish/bearish verdict
• Slope panel showing each MA's slope %, direction, and current value
• Support & resistance panel
• Recent golden/death cross list
• Active signals panel showing triggered trend/slope/curve/pattern signals
• Professional candlestick chart with colored MA overlays, trend lines, and signal markers

Signal System

The system now registers 34 signals across 4 categories:

Category	Examples	Count
Trend Lines	trendline_break_up, trendline_bounce_up, channel_top_touch	6
Slope Dynamics	slope_accelerating_up, slope_decelerating_up, bullish/bearish divergence	6
Curve / MA	ma_golden_cross, ma_convergence, ma_divergence, bb_squeeze	8
Candlestick Patterns	morning_star, bearish_engulfing, hammer, three_white_soldiers	10
Enhanced Classic	big_bull_breakout, support_bounce, reversal_after_decline	4

Each signal includes direction (bullish/bearish/neutral) and strength score, and contributes to a composite verdict.

Strategy Engine (NEW)

The system now treats every indicator "line" as a factor and combines multiple factors into actionable strategies.

• 20 strategies across 5 categories: Trend Following, Momentum Reversal, Volume Confirmation, Volatility Breakout, Multi-Confirmation
• Each strategy emits BUY / STRONG_BUY / SELL / STRONG_SELL signals
• Built-in backtest engine with stop-loss, take-profit, and max-hold-days
• Strategy ranking by Sharpe ratio, return, win rate, max drawdown, profit factor
• Multi-strategy consensus aggregates all 20 strategies into a single score

Strategy Portfolio (NEW)

Combines multiple strategies into a portfolio and adapts to market regimes:

• Market Regime Detection: 强趋势上涨 / 强趋势下跌 / 震荡市 / 顶部反转 / 底部反转
• Portfolio Methods:
  • Equal-weight top-5 strategies
  • Sharpe-weighted top-5 strategies
  • Regime-driven strategy selection
  • Dynamic volatility targeting
• Portfolio backtest with equity curve vs buy-and-hold benchmark
• Position sizing: Kelly ratio, volatility targeting, max drawdown controls

Example CLI output:

$ python src/analyzer.py META --period 6mo

[4.5/5] 策略组合分析...
  🌤 市场状态: 强趋势下跌 (置信度 100)
  📊 组合方式对比 (vs 买入持有 -12.8%):
    等权 Top5          总收益 14.1%  夏普 1.62  最大回撤 -9.0%  胜率 67%  交易6
    夏普加权 Top5        总收益 13.5%  夏普 1.58  最大回撤 -7.8%  胜率 67%  交易6
    动态波动率            总收益 11.3%  夏普 1.62  最大回撤 -7.3%  胜率 67%  交易6

$ python src/analyzer.py AAPL --period 6mo

🏆 Top 5 Strategies by Sharpe:
  1. Stochastic双线交叉   收益 11.4%  夏普 2.06  胜率 67%  交易3次
  2. OBV背离             收益  8.5%  夏普 1.06  胜率 75%  交易4次
  3. 均线多头排列         收益  5.7%  夏普 1.03  胜率100%  交易2次

🗳 多策略共识: 偏空 (评分: -2)

Web Dashboard

python src/api.py

Interactive dashboard with:

• Stock ticker input + period selector
• Overall trend card (上升 / 下降 / 横盘)
• MA alignment card (多头排列 / 空头排列 / 交织)
• Signal score card with bullish/bearish verdict
• Strategy panel showing backtest leaderboard + multi-strategy consensus
• Strategy signal chart showing BUY/SELL markers for top-performing strategies
• Slope panel, support/resistance, pattern search, factor table
• Professional candlestick charts with 30+ indicator lines and multi-panel mode

Project Structure

flow/
├── requirements.txt
├── README.md
├── KLINE_STRATEGY_FRAMEWORK.md
├── data/
│   ├── sector_etf_prices.csv
│   ├── event_sector_mapping.json
│   └── vix_data.csv
├── src/
│   ├── analyzer.py          # CLI entry point
│   ├── api.py               # Flask API + dashboard server
│   ├── data_loader.py       # yfinance / local CSV data loading
│   ├── indicators.py        # MA, slope, trend, crosses, support/resistance
│   ├── patterns.py          # Trend lines, slope, curves, candlestick patterns
│   ├── factors.py           # 85+ factor calculation (MA, volatility, momentum, volume, structure, adaptive)
│   ├── pattern_search.py    # Historical / cross-stock pattern similarity search
│   ├── strategies.py        # Multi-factor strategy engine + backtest framework
│   ├── stats.py             # Signal forward-return statistics
│   ├── report.py            # Markdown report generation
│   └── chart.py             # mplfinance candlestick charts with multi-panel support
└── web/
    └── index.html           # Dashboard frontend

Dependencies

• Python >= 3.10
• pandas, numpy, scipy
• matplotlib, mplfinance
• yfinance
• flask

License

MIT