Skip to content

Technical Indicators

Quantex ships with a growing library of pure-Pandas indicator functions that you can mix-and-match inside any strategy.

1. Quick-start

from quantex.indicators import sma, rsi
import pandas as pd

# DataFrame with a `close` column (can be price history or your own)
df = pd.read_csv("prices.csv")

# Add two new columns in one line each
# 20-bar Simple Moving Average
# 14-bar Relative Strength Index

df["sma_20"] = sma(df["close"], period=20)
df["rsi_14"] = rsi(df["close"], period=14)

All indicator functions preserve the original index and never mutate your input series.

2. Inside a Strategy

The Strategy base class exposes a helper method indicator() that computes and caches values automatically.

from quantex.strategy import Strategy

class SMACross(Strategy):
    fast = 10
    slow = 30

    def run(self):
        if self.index < self.slow:  # warm-up guard
            return

        fast_val = self.indicator("sma", "AAPL", period=self.fast)
        slow_val = self.indicator("sma", "AAPL", period=self.slow)

        price = self.get_price("AAPL")

        # Golden cross
        if fast_val > slow_val and self.positions["AAPL"].is_closed:
            self.buy("AAPL")
        # Death cross
        elif fast_val < slow_val and self.positions["AAPL"].is_long:
            self.sell("AAPL")

Behind the scenes Quantex:

  1. Looks up the sma function in quantex.indicators.
  2. Computes the entire series once the first time you call it.
  3. Stores the result in a per-strategy cache keyed by (name, symbol, params).
  4. Returns only the latest value for the current bar, so you can work with scalars.

This means subsequent calls during the simulation are O(1) regardless of look-back length.

3. Indicator palette (MVP)

Category Function Signature
Trend sma sma(series, period=20)
ema ema(series, period=20, adjust=False)
Momentum rsi rsi(series, period=14)
Volatility bollinger_bands bollinger_bands(series, period=20, std_dev=2)

Each function lives at quantex.indicators.<name>.

Future releases will add MACD, ATR, Stochastic, CCI, VWAP, Donchian channels, and more. Custom indicators can be added by simply creating a function that accepts a pd.Series and returns a pd.Series; place it in your own module and import it in your strategy.

4. Mermaid overview

graph TD EventBus("EventBus → price row, timeline") --> Strategy Strategy -->|calls| IndicatorCache IndicatorCache --> Indicators("quantex.indicators.*") Indicators -->|compute| Pandas/NumPy IndicatorCache --> Strategy

5. Performance tips

  • Prefer numeric NumPy dtypes (float64) – object dtype slows down rolling ops.
  • Don’t over-request: if you only need the latest value use indicator(); computing the full Series repeatedly is wasteful.
  • For very long histories (>5 M rows) consider installing TA-Lib and swapping implementations; Quantex will detect and use vectorised C routines automatically in future versions.

6. Reference API

All indicator functions follow this pattern:

def indicator_name(series: pd.Series, *args, **kwargs) -> pd.Series:
    """Docs with parameter explanation."""

Error handling

  • A ValueError is raised for invalid periods or parameters (e.g. period <= 0).
  • If not enough history is available the result contains NaN until the window is filled.

Read the docstrings (help(quantex.indicators.sma)) for full details.