indicators.py

"""
indicators.py
Plain pandas/numpy implementations — no TA-Lib dependency, so it installs
cleanly on Windows without a compiler toolchain.
"""
import numpy as np
import pandas as pd


def ema(series: pd.Series, period: int) -> pd.Series:
    return series.ewm(span=period, adjust=False).mean()


def sma(series: pd.Series, period: int) -> pd.Series:
    return series.rolling(period).mean()


def rsi(series: pd.Series, period: int = 14) -> pd.Series:
    delta = series.diff()
    gain = delta.clip(lower=0)
    loss = -delta.clip(upper=0)
    avg_gain = gain.ewm(alpha=1 / period, adjust=False).mean()
    avg_loss = loss.ewm(alpha=1 / period, adjust=False).mean()
    rs = avg_gain / avg_loss.replace(0, np.nan)
    out = 100 - (100 / (1 + rs))
    return out.fillna(50)


def atr(df: pd.DataFrame, period: int = 14) -> pd.Series:
    high, low, close = df["high"], df["low"], df["close"]
    prev_close = close.shift(1)
    tr = pd.concat(
        [(high - low), (high - prev_close).abs(), (low - prev_close).abs()], axis=1
    ).max(axis=1)
    return tr.ewm(alpha=1 / period, adjust=False).mean()


def macd(series: pd.Series, fast: int = 12, slow: int = 26, signal: int = 9):
    fast_ema = ema(series, fast)
    slow_ema = ema(series, slow)
    macd_line = fast_ema - slow_ema
    signal_line = ema(macd_line, signal)
    hist = macd_line - signal_line
    return macd_line, signal_line, hist


def bollinger_bands(series: pd.Series, period: int = 20, std_mult: float = 2.0):
    mid = sma(series, period)
    std = series.rolling(period).std()
    upper = mid + std_mult * std
    lower = mid - std_mult * std
    return upper, mid, lower


def stochastic_oscillator(df: pd.DataFrame, k_period: int = 8, k_slowing: int = 5, d_period: int = 3):
    """'Slow stochastic', the common retail convention for three numbers
    like (8, 5, 3): a raw %K over `k_period` bars, smoothed by a
    `k_slowing`-period SMA to produce the displayed %K line, then %D is
    a `d_period`-period SMA of that already-slowed %K. Note some
    platforms label these three inputs in a different order (e.g. MT5's
    dialog is %K period / %D period / Slowing) — if your source used a
    different platform's convention, the numbers may need reordering to
    match; this function documents exactly which role each argument
    plays so that's easy to check."""
    high, low, close = df["high"], df["low"], df["close"]
    lowest_low = low.rolling(k_period).min()
    highest_high = high.rolling(k_period).max()
    raw_k = 100 * (close - lowest_low) / (highest_high - lowest_low).replace(0, np.nan)
    slow_k = raw_k.rolling(k_slowing).mean()
    d = slow_k.rolling(d_period).mean()
    return slow_k.fillna(50), d.fillna(50)


def add_all_indicators(df: pd.DataFrame) -> pd.DataFrame:
    """Attaches a standard indicator set used by strategies.py and rating.py."""
    out = df.copy()
    out["ema_20"] = ema(out["close"], 20)
    out["ema_50"] = ema(out["close"], 50)
    out["ema_200"] = ema(out["close"], 200)
    out["rsi_14"] = rsi(out["close"], 14)
    out["atr_14"] = atr(out, 14)
    macd_line, signal_line, hist = macd(out["close"])
    out["macd"] = macd_line
    out["macd_signal"] = signal_line
    out["macd_hist"] = hist
    bb_up, bb_mid, bb_low = bollinger_bands(out["close"])
    out["bb_upper"] = bb_up
    out["bb_mid"] = bb_mid
    out["bb_lower"] = bb_low
    stoch_k, stoch_d = stochastic_oscillator(out, 8, 5, 3)
    out["stoch_k"] = stoch_k
    out["stoch_d"] = stoch_d
    out["body"] = (out["close"] - out["open"]).abs()
    out["range"] = out["high"] - out["low"]
    out["upper_wick"] = out["high"] - out[["close", "open"]].max(axis=1)
    out["lower_wick"] = out[["close", "open"]].min(axis=1) - out["low"]
    return out