pip install finkit

import pandas as pd
import finkit

# Technical indicators
prices = pd.Series([...])  # your price data
rsi = finkit.rsi(prices, period=14)
sma = finkit.sma(prices, period=20)
upper, middle, lower = finkit.bollinger_bands(prices)

# Risk metrics
returns = prices.pct_change().dropna()
sharpe = finkit.sharpe_ratio(returns)
sortino = finkit.sortino_ratio(returns)
mdd = finkit.max_drawdown(returns)
var_95 = finkit.var(returns, confidence=0.95, method="parametric")

from finkit import sma, ema, rsi, macd, bollinger_bands, atr, vwap, obv

# Moving averages
sma_20 = sma(prices, period=20)
ema_12 = ema(prices, period=12)

# Momentum
rsi_14 = rsi(prices, period=14)
macd_line, signal_line, histogram = macd(prices)

# Volatility
upper, middle, lower = bollinger_bands(prices, period=20, std_dev=2)
atr_14 = atr(high, low, close, period=14)

# Volume
vwap_values = vwap(high, low, close, volume)
obv_values = obv(close, volume)

from finkit import sharpe_ratio, sortino_ratio, max_drawdown, var

returns = prices.pct_change().dropna()

# Annualized Sharpe (assumes 252 trading days)
sharpe = sharpe_ratio(returns, risk_free_rate=0.05)

# Sortino (downside deviation only)
sortino = sortino_ratio(returns)

# Maximum drawdown
mdd = max_drawdown(returns)  # returns negative float, e.g. -0.15

# Value at Risk
var_hist = var(returns, confidence=0.95, method="historical")
var_param = var(returns, confidence=0.99, method="parametric")

from finkit import Rule, screen

universe = pd.DataFrame({
    "symbol": ["AAPL", "GOOGL", "MSFT", "AMZN"],
    "pe_ratio": [28, 22, 35, 45],
    "market_cap": [3e12, 2e12, 3.1e12, 1.9e12],
    "sector": ["Tech", "Tech", "Tech", "Consumer"],
})

results = screen(universe, rules=[
    Rule("pe_ratio", "<", 30),
    Rule("market_cap", ">", 2e12),
    Rule("sector", "==", "Tech"),
])
# Returns: AAPL, GOOGL

from finkit import crossover, crossunder, divergence, SignalEngine

# Crossover detection
buy_signals = crossover(fast=ema_12, slow=sma_20)
sell_signals = crossunder(fast=ema_12, slow=sma_20)

# Composable signal engine
engine = SignalEngine()
engine.add_rule("rsi_oversold", lambda df: df["rsi"] < 30, direction="long")
engine.add_rule("macd_cross", lambda df: crossover(df["macd"], df["signal"]), direction="long")
engine.add_rule("rsi_overbought", lambda df: df["rsi"] > 70, direction="short")

signals = engine.evaluate(df)
# Returns DataFrame with: signal_score, direction, triggered_rules

from finkit import calculate_trade_cost, annual_cost_drag

cost = calculate_trade_cost(shares=100, price=150.0, commission_per_share=0.005)
# TradeCost(commission=0.50, total=0.50, cost_bps=0.33)

drag = annual_cost_drag(trades_per_year=50, avg_cost_bps=0.33)
# 16.5 bps annual drag

from vectrade import VecTrade
import finkit

client = VecTrade()

# Fetch historical prices
candles = client.candles.get("AAPL", interval="1d", limit=100)
prices = pd.Series([c.close for c in candles])

# Local computation (no API calls)
rsi_values = finkit.rsi(prices)
sharpe = finkit.sharpe_ratio(prices.pct_change().dropna())