motosan-ta-sdk

Core Python SDK for technical analysis — 20 indicators, pattern detection, screening, and risk assessment. Strict parity with pandas-ta reference vectors.

• GitHub: daiwanwei/motosan-ta
• Changelog: CHANGELOG.md
• Rust crates: motosan-ta-math, motosan-ta-stream

Installation

pip install motosan-ta-sdk

# With disk caching support
pip install motosan-ta-sdk[cache]

Quick Start

from ta_sdk import YFinanceProvider, sma, ema, rsi, macd, atr, check_stage2, detect_vcp

provider = YFinanceProvider()
ohlcv = provider.get_ohlcv("AAPL", period="1y", interval="1d")
df = ohlcv.df

# Indicators
sma_50 = sma(df["close"], 50)
ema_21 = ema(df["close"], 21)
rsi_14 = rsi(df["close"], 14)
macd_line, signal, hist = macd(df["close"])
atr_14 = atr(df["high"], df["low"], df["close"], 14)

# Screen for Stage 2 trend
result = check_stage2(df)
print(f"Stage 2: {result['passes']} ({result['score']}/{result['max_score']})")

# Detect VCP pattern
vcp = detect_vcp(df)
print(f"VCP: {vcp.detected} (confidence: {vcp.confidence:.0%})")

Indicators (20 total)

Moving Averages (7): SMA, EMA, KAMA, DEMA, TEMA, HMA, SMMA Momentum (10): RSI, MACD, Stochastic, CCI, Williams %R, MFI, ROC, Aroon, TSI, Fisher Transform, TRIX Volatility (6): ATR, Bollinger Bands, ADX, Keltner Channels, Donchian Channels, Supertrend Volume (2): OBV, CMF

All indicators use SMA-seeded EMA / Wilder's RMA to match pandas-ta output.

Top-Level Exports

from ta_sdk import (
    # Data
    OHLCVData, Quote, PatternResult, AnalysisReport,
    YFinanceProvider, BinanceProvider, MultiProvider,
    # Indicators — Moving Averages
    sma, ema, kama, dema, tema, hma, smma,
    # Indicators — Momentum
    rsi, macd, aroon, tsi, fisher_transform, trix,
    # Indicators — Volatility
    atr, bollinger_bands,
    # Indicators — Volume
    obv, cmf,
    # Screeners
    check_stage2, check_entry_signal,
    # Patterns
    detect_vcp, detect_cup_handle, detect_double_bottom,
    # Analysis
    analyze_stock,
)

Additional indicators available via direct module import:

from ta_sdk.indicators.momentum import stochastic, cci, roc, williams_r, mfi
from ta_sdk.indicators.volatility import adx, keltner, donchian, supertrend

Module Overview

Data Providers

MultiProvider

Routes symbols to the correct provider automatically based on symbol format:

from ta_sdk import YFinanceProvider, BinanceProvider, MultiProvider
from ta_sdk.data.models import Market

multi = MultiProvider()
multi.register(Market.US, YFinanceProvider())
multi.register(Market.TW, YFinanceProvider())
multi.register(Market.CRYPTO, BinanceProvider())

multi.get_ohlcv("AAPL")      # Auto-detected as US
multi.get_ohlcv("2330.TW")   # Auto-detected as TW
multi.get_ohlcv("BTCUSDT")   # Auto-detected as Crypto

CachedProvider

Wraps any DataProvider with disk-based caching. Requires the [cache] extra.

from ta_sdk.data.cached_provider import CachedProvider

cached = CachedProvider(
    provider=YFinanceProvider(),
    cache_dir=".ta_cache",
    ohlcv_ttl=3600,   # 1 hour
    quote_ttl=60,      # 1 minute
)
ohlcv = cached.get_ohlcv("AAPL")

Supported Markets

Requirements

• Python >= 3.12
• pandas >= 2.2
• numpy >= 1.26
• yfinance >= 0.2.36
• python-binance >= 1.0.19
• requests >= 2.31
• diskcache >= 5.6 (optional, for [cache] extra)

License

MIT