techkit 1.2.1

High-performance technical analysis library with 153 indicators, 100% TA-Lib compatible

TechKit

TechKit is a high-performance technical analysis library for Python, featuring:

• 🚀 153 Technical Indicators - Complete coverage of all TA-Lib indicators
• ⚡ Blazing Fast - C++ core with O(1) incremental updates
• 🔄 100% TA-Lib Compatible - Drop-in replacement API
• 📊 61 Candlestick Patterns - Full CDL pattern recognition
• 🧵 Thread-Safe - Zero global state, parallel-ready
• 📦 Zero Dependencies - Only NumPy required

Installation

pip install techkit

Platform Support

Platform	Python	Architecture
Linux	3.10-3.13	x86_64
macOS	3.10-3.13	x86_64, arm64
Windows	3.10-3.13	AMD64

Quick Start

Basic Usage

import numpy as np
from techkit import SMA, RSI, MACD, BBANDS

# Sample price data
prices = np.random.randn(100).cumsum() + 100

# Create indicators
sma = SMA(period=20)
rsi = RSI(period=14)
macd = MACD(fast=12, slow=26, signal=9)

# Calculate (batch)
sma_values = sma.calculate(prices)
rsi_values = rsi.calculate(prices)
macd_result = macd.calculate(prices)

print(f"SMA: {sma_values[-1]:.2f}")
print(f"RSI: {rsi_values[-1]:.2f}")
print(f"MACD: {macd_result.macd[-1]:.4f}")

Streaming / Incremental Updates

from techkit import SMA, RSI

sma = SMA(20)
rsi = RSI(14)

# Process bar by bar (O(1) per update)
for price in live_price_stream():
    sma_result = sma.update(price)
    rsi_result = rsi.update(price)
    
    if sma_result.valid and rsi_result.valid:
        print(f"SMA: {sma_result.value:.2f}, RSI: {rsi_result.value:.2f}")

TA-Lib Drop-in Replacement

# Before (TA-Lib)
import talib
sma = talib.SMA(prices, timeperiod=20)
rsi = talib.RSI(prices, timeperiod=14)
macd, signal, hist = talib.MACD(prices)

# After (TechKit) - Same API!
from techkit import talib_compat as ta
sma = ta.SMA(prices, timeperiod=20)
rsi = ta.RSI(prices, timeperiod=14)
macd, signal, hist = ta.MACD(prices)

Candlestick Patterns

from techkit import CDL_HAMMER, CDL_ENGULFING, CDL_MORNINGSTAR

# Detect patterns
hammer = CDL_HAMMER()
signals = hammer.calculate(open, high, low, close)

# Signals: +100 (bullish), -100 (bearish), 0 (none)
bullish_hammers = np.where(signals > 0)[0]

Indicator Chaining

from techkit import Chain, RSI, EMA, SMA

# Smoothed RSI: RSI(14) → EMA(9)
smoothed_rsi = Chain([RSI(14), EMA(9)])
result = smoothed_rsi.calculate(prices)

# Double smoothed: SMA(10) → SMA(5)
double_smooth = Chain([SMA(10), SMA(5)])
result = double_smooth.calculate(prices)

Pandas Integration

import pandas as pd
from techkit import talib_compat as ta

df = pd.read_csv('ohlcv.csv')

# Directly use DataFrame columns
df['SMA_20'] = ta.SMA(df['close'], timeperiod=20)
df['RSI_14'] = ta.RSI(df['close'], timeperiod=14)
df['ATR_14'] = ta.ATR(df['high'], df['low'], df['close'], timeperiod=14)

macd, signal, hist = ta.MACD(df['close'])
df['MACD'] = macd
df['MACD_Signal'] = signal
df['MACD_Hist'] = hist

Indicator Categories

Moving Averages (9)

SMA, EMA, WMA, DEMA, TEMA, KAMA, TRIMA, T3, MA

Momentum Indicators (26)

RSI, MACD, STOCH, STOCHF, STOCHRSI, CCI, ADX, ADXR, MOM, ROC, ROCP, ROCR, ROCR100, WILLR, MFI, ULTOSC, TRIX, APO, PPO, CMO, DX, PLUS_DI, MINUS_DI, PLUS_DM, MINUS_DM, AROON, AROONOSC, BOP

Volatility (4)

ATR, NATR, BBANDS, TRANGE

Volume (4)

OBV, AD, ADOSC, MFI

Statistics (9)

STDDEV, VAR, LINEARREG, LINEARREG_SLOPE, LINEARREG_INTERCEPT, LINEARREG_ANGLE, TSF, BETA, CORREL

Price Transform (4)

AVGPRICE, MEDPRICE, TYPPRICE, WCLPRICE

Math (6)

MAX, MIN, SUM, MIDPOINT, MIDPRICE, MINMAX

Cycle / Hilbert Transform (6)

HT_DCPERIOD, HT_DCPHASE, HT_TRENDMODE, HT_TRENDLINE, HT_PHASOR, HT_SINE

Parabolic SAR (2)

SAR, SAREXT

Candlestick Patterns (61)

CDL_DOJI, CDL_HAMMER, CDL_ENGULFING, CDL_MORNINGSTAR, CDL_EVENINGSTAR, CDL_3WHITESOLDIERS, CDL_3BLACKCROWS, CDL_HARAMI, CDL_PIERCING, CDL_DARKCLOUDCOVER, and 51 more...

Advanced Analytics (17)

Risk Metrics: SharpeRatio, SortinoRatio, CalmarRatio, MaxDrawdown, HistoricalVaR, CVaR

Volatility Models: EWMAVolatility, RealizedVolatility, ParkinsonVolatility, GARCHVolatility

Structure Analysis: ZigZag, SwingHighLow, PivotPoints

Pattern Recognition: HarmonicPattern, ChartPattern

Performance

TechKit uses C++ with O(1) incremental computation:

Operation	TechKit	TA-Lib
Single update	O(1)	O(n)
Batch (1000 bars)	~0.1ms	~0.2ms
Memory per indicator	O(period)	O(n)

Benchmark Example

import time
import numpy as np
from techkit import SMA

prices = np.random.randn(1_000_000).cumsum() + 100

# Batch mode
start = time.time()
result = SMA(20).calculate(prices)
print(f"Batch (1M points): {time.time() - start:.3f}s")

# Streaming mode
sma = SMA(20)
start = time.time()
for price in prices[:100_000]:
    sma.update(price)
print(f"Streaming (100K points): {time.time() - start:.3f}s")

Version Information

import techkit

print(techkit.__version__)       # "1.2.1"
print(techkit.version())         # "1.2.1"
print(techkit.version_tuple())   # (1, 2, 1)

License

MIT License - see LICENSE for details.

Links

• GitHub Repository
• Issue Tracker
• Changelog