/* tslint:disable */
/* eslint-disable */

/* auto-generated by NAPI-RS */

export interface RenkoBrick {
  price: number
  direction: string
}
export declare function renkoChart(prices: Array<number>, brickSize?: number): Array<RenkoBrick>
export interface KagiPoint {
  price: number
  direction: string
}
export declare function kagiChart(prices: Array<number>, reversalAmount?: number): Array<KagiPoint>
export declare function lowHighOpenCloseVolumeDateToArray(data: Array<MarketData>): MarketDataResult
export interface BollingerBandsResult {
  middle: Array<number>
  upper: Array<number>
  lower: Array<number>
}
export interface MarketData {
  low: number
  high: number
  open: number
  close: number
  volume: number
  date: string
}
export interface MarketDataResult {
  lows: Array<number>
  highs: Array<number>
  opens: Array<number>
  closes: Array<number>
  volumes: Array<number>
  dates: Array<string>
}
export interface RegressionSegment {
  /** Start index in the original data */
  startIndex: number
  /** End index in the original data */
  endIndex: number
  /** Slope (trend direction and strength) */
  slope: number
  /** Intercept */
  intercept: number
  /** Residual standard deviation */
  stdDev: number
  /** Fitted values for this segment */
  fitted: Array<number>
  /** Upper band (fitted + mult * std_dev) */
  upperBand: Array<number>
  /** Lower band (fitted - mult * std_dev) */
  lowerBand: Array<number>
}
export interface AnchoredRegressionResult {
  /** All regression segments */
  segments: Array<RegressionSegment>
  /** Full-length fitted line (NaN where no regression) */
  fitted: Array<number>
  /** Full-length upper band */
  upperBand: Array<number>
  /** Full-length lower band */
  lowerBand: Array<number>
  /** Full-length slope values (slope of the segment at each bar) */
  slopes: Array<number>
}
/**
 * Static Anchored Regression
 *
 * Divides the price series into fixed segments based on `anchor_period` bars.
 * Each segment gets its own independent linear regression.
 *
 * Parameters:
 * - prices: closing prices
 * - anchor_period: number of bars per segment (e.g. 5 for weekly on daily data)
 * - band_mult: multiplier for std dev bands (default: 1.0)
 */
export declare function anchoredRegressionStatic(prices: Array<number>, anchorPeriod: number, bandMult?: number | undefined | null): AnchoredRegressionResult
/**
 * Rolling Anchored Regression
 *
 * Regression updates bar-by-bar from each anchor reset point.
 * The anchor resets every `anchor_period` bars.
 * At each bar, the regression is computed from the last anchor point to the current bar.
 *
 * Parameters:
 * - prices: closing prices
 * - anchor_period: bars between anchor resets (e.g. 5 for weekly on daily data)
 * - band_mult: multiplier for std dev bands (default: 1.0)
 */
export declare function anchoredRegressionRolling(prices: Array<number>, anchorPeriod: number, bandMult?: number | undefined | null): AnchoredRegressionResult
export interface AwesomeOscillatorResult {
  /** AO values (SMA5 - SMA34 of midpoints) */
  ao: Array<number>
  /** AO histogram color: 1 = green (rising), -1 = red (falling), 0 = neutral */
  histogram: Array<number>
}
/**
 * Awesome Oscillator (Bill Williams)
 *
 * AO = SMA(5, Midpoint) - SMA(34, Midpoint)
 * where Midpoint = (High + Low) / 2
 *
 * Measures market momentum. Histogram bars are green when AO is rising,
 * red when falling. Zero-line crossovers signal trend changes.
 *
 * Parameters:
 * - data: OHLCV market data
 * - fast_period: fast SMA period (default: 5)
 * - slow_period: slow SMA period (default: 34)
 */
export declare function awesomeOscillator(data: Array<MarketData>, fastPeriod?: number | undefined | null, slowPeriod?: number | undefined | null): AwesomeOscillatorResult
export declare function bollingerBands(data: Array<number>, period?: number | undefined | null, multiplier?: number | undefined | null): BollingerBandsResult
export interface CandlestickPatterns {
  /** Doji: +1 detected, 0 none */
  doji: Array<number>
  /** Bullish Engulfing: +1, Bearish Engulfing: -1 */
  engulfing: Array<number>
  /** Hammer: +1 (bullish reversal signal) */
  hammer: Array<number>
  /** Hanging Man: -1 (bearish reversal signal) */
  hangingMan: Array<number>
  /** Bullish Harami: +1, Bearish Harami: -1 */
  harami: Array<number>
  /** Morning Star: +1 (bullish three-bar reversal) */
  morningStar: Array<number>
  /** Evening Star: -1 (bearish three-bar reversal) */
  eveningStar: Array<number>
  /** Three White Soldiers: +1 (strong bullish) */
  threeWhiteSoldiers: Array<number>
  /** Three Black Crows: -1 (strong bearish) */
  threeBlackCrows: Array<number>
  /** Shooting Star: -1 (bearish reversal) */
  shootingStar: Array<number>
  /** Inverted Hammer: +1 (potential bullish reversal) */
  invertedHammer: Array<number>
  /** Spinning Top: +1 (indecision) */
  spinningTop: Array<number>
  /** Marubozu: +1 bullish (no shadows), -1 bearish */
  marubozu: Array<number>
  /** Composite signal: sum of all pattern signals at each bar */
  composite: Array<number>
}
/**
 * Detect common candlestick patterns from OHLC data.
 *
 * Returns +1 for bullish patterns, -1 for bearish, 0 for none.
 * All 13 patterns are computed in a single pass for efficiency.
 *
 * Parameters:
 * - data: OHLCV market data
 * - body_threshold: max body/range ratio for doji (default: 0.05 = 5%)
 */
export declare function candlestickPatterns(data: Array<MarketData>, bodyThreshold?: number | undefined | null): CandlestickPatterns
export interface ChoppinessResult {
  /** Choppiness Index values (0-100). NaN for warmup period. */
  chop: Array<number>
  /**
   * Signals: 1 = trending (CHOP crosses below low_threshold),
   * -1 = choppy/ranging (CHOP crosses above high_threshold), 0 = neutral
   */
  signals: Array<number>
}
/**
 * Choppiness Index (CI)
 *
 * CI = 100 * log10(Sum(TR, N) / (HighestHigh_N - LowestLow_N)) / log10(N)
 *
 * Measures whether the market is trending or range-bound:
 * - Low values (< 38.2) indicate a strong trend
 * - High values (> 61.8) indicate a choppy/sideways market
 *
 * Parameters:
 * - data: OHLCV market data
 * - period: lookback period (default: 14)
 * - low_threshold: below this = trending (default: 38.2)
 * - high_threshold: above this = choppy (default: 61.8)
 */
export declare function choppinessIndex(data: Array<MarketData>, period?: number | undefined | null, lowThreshold?: number | undefined | null, highThreshold?: number | undefined | null): ChoppinessResult
export interface ConditionalProbabilityResult {
  /** Probability of a move >= +second_threshold after a first move >= first_threshold */
  upProbability: number
  /** Probability of a move <= -second_threshold after a first move >= first_threshold */
  downProbability: number
  /** Number of times the first move condition was met */
  firstMoveCount: number
  /** Number of times the second move was up after first move */
  upCount: number
  /** Number of times the second move was down after first move */
  downCount: number
  /** Indices where up moves occurred (in original data) */
  upIndices: Array<number>
  /** Indices where down moves occurred (in original data) */
  downIndices: Array<number>
  /** All second move percentage changes (when first condition was met) */
  secondMoveReturns: Array<number>
}
/**
 * Conditional Probability Analysis
 *
 * Calculates: P(second_move >= threshold | first_move >= threshold)
 *
 * Given a price series, finds all instances where the price moved by at least
 * `first_threshold` over `first_move_days`, then measures what happened over
 * the following `second_move_days`.
 *
 * The first move is triggered by absolute change >= first_threshold (both up and down).
 * The second move probabilities are split into up (>= second_threshold) and down (<= -second_threshold).
 */
export declare function conditionalProbability(prices: Array<number>, firstMoveDays: number, secondMoveDays: number, firstThreshold: number, secondThreshold: number): ConditionalProbabilityResult
export interface ConditionalMatrixEntry {
  firstThreshold: number
  secondThreshold: number
  upProbability: number
  downProbability: number
  sampleCount: number
}
/**
 * Compute a matrix of conditional probabilities across multiple threshold combinations.
 *
 * Useful for heatmap visualization: for each (first_threshold, second_threshold) pair,
 * returns the up and down probabilities.
 */
export declare function conditionalProbabilityMatrix(prices: Array<number>, firstMoveDays: number, secondMoveDays: number, firstThresholds: Array<number>, secondThresholds: Array<number>): Array<ConditionalMatrixEntry>
export interface CopulaSample {
  u: Array<number>
  v: Array<number>
}
export interface CopulaFitResult {
  copulaType: string
  parameter: number
  logLikelihood: number
}
export interface ScenarioResult {
  ticker: string
  meanReturn: number
  worstCase: number
  bestCase: number
  simulatedReturns: Array<number>
}
export declare function quantileTransform(data: Array<number>): Array<number>
export declare function gaussianCopulaSample(rho: number, nSamples: number, seed?: number | undefined | null): CopulaSample
export declare function gaussianConditionalSample(u1: number, rho: number, nSamples: number, seed?: number | undefined | null): CopulaSample
export declare function claytonCopulaSample(theta: number, nSamples: number, seed?: number | undefined | null): CopulaSample
export declare function gumbelCopulaSample(theta: number, nSamples: number, seed?: number | undefined | null): CopulaSample
export declare function frankCopulaSample(theta: number, nSamples: number, seed?: number | undefined | null): CopulaSample
export declare function fitCopula(u: Array<number>, v: Array<number>, copulaType: string): CopulaFitResult
export declare function portfolioScenario(returnsData: Array<Array<number>>, marketDrop: number, copulaType?: string | undefined | null, nSimulations?: number | undefined | null): Array<ScenarioResult>
export interface DmiResult {
  plusDi: Array<number>
  minusDi: Array<number>
  adx: Array<number>
}
export declare function directionalMovementIndex(data: Array<MarketData>, period: number): DmiResult
/**
 * Disparity Index
 *
 * Measures the percentage distance between the current price and a moving average.
 * DI = 100 * (Close - MA(N)) / MA(N)
 *
 * Positive values: price is above the MA (bullish)
 * Negative values: price is below the MA (bearish)
 * Extreme values suggest overbought/oversold conditions
 */
export declare function disparityIndex(prices: Array<number>, period?: number | undefined | null): Array<number>
export interface Signal {
  type: number
  price: number
  index: number
}
export declare function entryExitSignals(data: Array<MarketData>, smaPeriod: number, emaPeriod: number, atrPeriod: number, threshold: number): Array<Signal>
export declare function exponentialMovingAverage(data: Array<number>, period: number): Array<number>
export interface ImportantLevels {
  highestResistance: number
  lowestSupport: number
  averagePivot: number
  supports: Array<number>
  resistances: Array<number>
}
export declare function extractImportantLevels(data: Array<number>): ImportantLevels
export interface FeatureRow {
  /** Bar index */
  index: number
  /** 1-bar return (pct change) */
  return1: number
  /** 5-bar return */
  return5: number
  /** 10-bar return */
  return10: number
  /** 20-bar return */
  return20: number
  /** True Range */
  trueRange: number
  /** ATR (Wilder's, 14-period) */
  atr14: number
  /** Rolling std dev of 1-bar returns (20-period) */
  volatility20: number
  /** High-Low range as % of close */
  rangePct: number
  /** RSI (14-period, Wilder's) */
  rsi14: number
  /** Rate of Change (10-period) */
  roc10: number
  /** Momentum (close - close[10]) */
  momentum10: number
  /** SMA 5 */
  sma5: number
  /** SMA 20 */
  sma20: number
  /** SMA 50 */
  sma50: number
  /** EMA 12 */
  ema12: number
  /** EMA 26 */
  ema26: number
  /** MACD line (EMA12 - EMA26) */
  macd: number
  /** MACD signal (EMA9 of MACD) */
  macdSignal: number
  /** MACD histogram */
  macdHistogram: number
  /** Bollinger %B: (close - lower) / (upper - lower) */
  bbPctB: number
  /** Bollinger bandwidth: (upper - lower) / middle */
  bbBandwidth: number
  /** Close relative to SMA20: (close - sma20) / sma20 */
  closeToSma20: number
  /** Close relative to SMA50 */
  closeToSma50: number
  /** Distance from 20-bar high (%) */
  distFromHigh20: number
  /** Distance from 20-bar low (%) */
  distFromLow20: number
  /** Volume change (pct) */
  volumeChange: number
  /** Volume / SMA20 of volume */
  volumeRatio: number
  /** Body size: |close - open| / (high - low) */
  bodyRatio: number
  /** Upper shadow: (high - max(open,close)) / (high - low) */
  upperShadow: number
  /** Lower shadow: (min(open,close) - low) / (high - low) */
  lowerShadow: number
  /** Gap: (open - prev_close) / prev_close */
  gap: number
  /** SMA5 > SMA20 (1.0 or 0.0) */
  trendSma520: number
  /** SMA20 > SMA50 (1.0 or 0.0) */
  trendSma2050: number
}
/**
 * Generate a complete feature matrix from OHLCV data for ML pipelines.
 *
 * Computes ~35 features per bar covering returns, volatility, momentum,
 * moving averages, MACD, Bollinger Bands, price position, volume,
 * candle patterns, and trend signals.
 *
 * First ~50 bars are skipped (warmup period). Returns one FeatureRow per valid bar.
 */
export declare function featureEngine(data: Array<MarketData>): Array<FeatureRow>
export interface FramaResult {
  /** FRAMA values (adaptive moving average) */
  frama: Array<number>
  /** Fractal dimension at each bar (1.0 = trending, 2.0 = choppy) */
  fractalDimension: Array<number>
  /** Alpha (smoothing factor) at each bar */
  alpha: Array<number>
  /** FRAMA slope (bar-to-bar change, for trend detection) */
  slope: Array<number>
}
/**
 * Fractal Adaptive Moving Average (FRAMA) — John Ehlers
 *
 * An EMA whose smoothing factor adapts based on the fractal dimension of prices.
 * In trending markets (fractal dim ~1): FRAMA is fast and responsive.
 * In choppy markets (fractal dim ~2): FRAMA is slow and smooth.
 *
 * Fractal dimension is estimated by comparing the price range over N bars
 * to the ranges of two N/2 sub-periods (Ehlers' method).
 *
 * Parameters:
 * - data: OHLCV market data
 * - period: lookback for fractal calculation (default: 20, must be even)
 * - fast_period: fast EMA equivalent period when trending (default: 4)
 * - slow_period: slow EMA equivalent period when choppy (default: 200)
 */
export declare function frama(data: Array<MarketData>, period?: number | undefined | null, fastPeriod?: number | undefined | null, slowPeriod?: number | undefined | null): FramaResult
export interface GmmCluster {
  /** Cluster index */
  id: number
  /** Mean of each feature dimension */
  mean: Array<number>
  /** Variance of each feature dimension (diagonal covariance) */
  variance: Array<number>
  /** Mixing weight (proportion of data in this cluster) */
  weight: number
  /** Number of points assigned to this cluster */
  count: number
}
export interface GmmResult {
  /** Cluster assignment for each data point */
  labels: Array<number>
  /**
   * Posterior probabilities: labels.len() * n_components, row-major
   * probabilities[i * n_components + k] = P(cluster k | point i)
   */
  probabilities: Array<number>
  /** Cluster details */
  clusters: Array<GmmCluster>
  /** BIC score (lower = better model fit vs complexity) */
  bic: number
  /** Log-likelihood of the fitted model */
  logLikelihood: number
  /** Number of EM iterations performed */
  iterations: number
}
/**
 * Gaussian Mixture Model (GMM) via Expectation-Maximization
 *
 * Clusters multi-dimensional data into n_components Gaussian distributions.
 * Useful for market regime detection (e.g., calm/volatile/transition states).
 *
 * Input: a flat array of features, row-major, with `n_features` per row.
 * Example: for returns + volume_change with 100 bars:
 *   data = [ret_0, vol_0, ret_1, vol_1, ...] with n_features = 2
 *
 * Parameters:
 * - data: flat array of feature values (row-major)
 * - n_features: number of features per observation
 * - n_components: number of Gaussian clusters (default: 3)
 * - max_iterations: max EM iterations (default: 100)
 * - tolerance: convergence threshold on log-likelihood change (default: 1e-6)
 * - normalize: if true, z-score normalize each feature before fitting (default: true)
 * - seed: optional random seed for reproducibility
 */
export declare function gaussianMixture(data: Array<number>, nFeatures: number, nComponents?: number | undefined | null, maxIterations?: number | undefined | null, tolerance?: number | undefined | null, normalize?: boolean | undefined | null, seed?: number | undefined | null): GmmResult
export interface HarResult {
  /** Predicted volatility at each bar (annualized) */
  predictedVol: Array<number>
  /** Daily volatility component (Yang-Zhang, 1-day) */
  volDaily: Array<number>
  /** Weekly volatility component (5-day average of daily vol) */
  volWeekly: Array<number>
  /** Monthly volatility component (22-day average of daily vol) */
  volMonthly: Array<number>
  /** Volatility regime: 0=low, 1=medium, 2=high, -1=warmup */
  regime: Array<number>
  /** Suggested exposure: 2.0 (low vol), 1.0 (medium), 0.0 (high) */
  exposure: Array<number>
}
/**
 * HAR-X Volatility Model (Heterogeneous Autoregressive with eXogenous variables)
 *
 * Combines volatility from multiple timeframes:
 * - Daily (1-day Yang-Zhang volatility)
 * - Weekly (5-day average)
 * - Monthly (22-day average)
 *
 * Predicts future volatility using: vol_pred = a + b1*vol_d + b2*vol_w + b3*vol_m + b4*vix
 * Coefficients estimated via rolling OLS regression.
 *
 * Regime classification based on rolling percentiles:
 * - Low vol (< percentile_low): exposure = 2.0 (leveraged)
 * - Medium vol: exposure = 1.0 (normal)
 * - High vol (> percentile_high): exposure = 0.0 (cash)
 *
 * Parameters:
 * - data: OHLCV market data
 * - yz_window: Yang-Zhang window (default: 10)
 * - har_lookback: OLS regression lookback (default: 252)
 * - percentile_low: low vol threshold (default: 25)
 * - percentile_high: high vol threshold (default: 75)
 * - vix_data: optional VIX values (same length as data) for HAR-X extension
 */
export declare function harVolatility(data: Array<MarketData>, yzWindow?: number | undefined | null, harLookback?: number | undefined | null, percentileLow?: number | undefined | null, percentileHigh?: number | undefined | null, vixData?: Array<number> | undefined | null): HarResult
export interface IchimokuData {
  tenkanSen: number
  kijunSen: number
  senkouSpanA: number
  senkouSpanB: number
  chikouSpan: number
}
export declare function ichimoku(data: Array<MarketData>, tenkanPeriod?: number, kijunPeriod?: number, senkouBPeriod?: number, chikouShift?: number): Array<IchimokuData>
export interface KReversalResult {
  kValues: Array<number>
  buySignals: Array<KReversalSignal>
  sellSignals: Array<KReversalSignal>
}
export interface KReversalSignal {
  index: number
  price: number
  kValue: number
}
/**
 * K-Reversal Indicator
 *
 * K = 100 * (Close - Low_N) / (High_N - Low_N)
 *
 * K < buy_threshold (default 20) suggests potential uptrend (oversold)
 * K > sell_threshold (default 80) suggests potential downtrend (overbought)
 */
export declare function kReversal(data: Array<MarketData>, period?: number | undefined | null, buyThreshold?: number | undefined | null, sellThreshold?: number | undefined | null): KReversalResult
export interface OptionContract {
  /** Strike price */
  strike: number
  /** Open interest */
  openInterest: number
  /** Daily volume */
  volume: number
  /** Days to expiry */
  dte: number
  /** "call" or "put" */
  side: string
  /** Implied volatility (optional, 0 if unknown) */
  impliedVolatility: number
}
export interface ScoredOption {
  /** Original contract index */
  index: number
  strike: number
  openInterest: number
  volume: number
  dte: number
  side: string
  impliedVolatility: number
  /** OI/Volume ratio (capped) */
  oiVolumeRatio: number
  /** Z-score of open interest within its expiry group */
  oiZScore: number
  /** Z-score of OI/Volume ratio within its expiry group */
  ovZScore: number
  /** Percentile rank of OI z-score (0-1) */
  oiPercentile: number
  /** Percentile rank of OV z-score (0-1) */
  ovPercentile: number
  /** OTM distance factor */
  otmFactor: number
  /** DTE decay factor */
  dteFactor: number
  /** Final composite score (higher = more institutional interest) */
  score: number
}
/**
 * Big Money Options Flow Scoring
 *
 * Ranks option contracts by institutional interest using a composite score:
 * Score = w_oi * OI_percentile * dte_factor + w_ov * OV_percentile * dte_factor + w_otm * otm_factor
 *
 * Parameters:
 * - contracts: array of option contracts with strike, OI, volume, DTE, side
 * - spot_price: current underlying price
 * - top_n: number of top contracts to return (default: 50)
 * - k_otm: OTM scaling factor (default: 2.0, higher = more penalty for far strikes)
 * - min_volume: minimum volume filter (default: 10)
 * - min_oi: minimum open interest filter (default: 100)
 * - cap_oi_vol: cap for OI/Volume ratio (default: 100)
 * - w_oi: weight on OI z-score percentile (default: 0.4)
 * - w_ov: weight on OI/Volume z-score percentile (default: 0.4)
 * - w_otm: weight on OTM distance (default: 0.2)
 */
export declare function optionsFlowScore(contracts: Array<OptionContract>, spotPrice: number, topN?: number | undefined | null, kOtm?: number | undefined | null, minVolume?: number | undefined | null, minOi?: number | undefined | null, capOiVol?: number | undefined | null, wOi?: number | undefined | null, wOv?: number | undefined | null, wOtm?: number | undefined | null): Array<ScoredOption>
export declare function parabolicSar(data: Array<MarketData>, start?: number | undefined | null, increment?: number | undefined | null, maxValue?: number | undefined | null): Array<number>
export interface PatternMemoryResult {
  /**
   * Directional signal at each bar: sum of labels from k-nearest neighbors.
   * Positive = historically bullish, negative = historically bearish.
   */
  signal: Array<number>
  /** Normalized signal: signal / k (range -1 to +1) */
  normalizedSignal: Array<number>
  /** Number of bullish neighbors at each bar */
  bullishCount: Array<number>
  /** Number of bearish neighbors at each bar */
  bearishCount: Array<number>
  /** Average Lorentzian distance to the k-nearest neighbors */
  avgDistance: Array<number>
}
/**
 * Pattern Memory (Lorentzian Classification)
 *
 * Non-parametric, memory-based directional signal. For each bar:
 * 1. Encode market state as a feature vector (5 indicators x window bars)
 * 2. Compare to all past states within a lookback using Lorentzian distance
 * 3. Find k-nearest neighbors and check what followed (+1 up, -1 down)
 * 4. Sum the labels as a directional signal
 *
 * Features computed internally:
 * - RSI(14), WaveTrend(10,11), CCI(20), ADX(14), RSI(9)
 *
 * Parameters:
 * - data: OHLCV market data
 * - k_neighbors: number of nearest neighbors (default: 100)
 * - lookback: how many past bars to search (default: 200)
 * - window: number of consecutive bars per feature vector (default: 5)
 * - forward_bars: bars ahead to determine label (default: 4)
 */
export declare function patternMemory(data: Array<MarketData>, kNeighbors?: number | undefined | null, lookback?: number | undefined | null, window?: number | undefined | null, forwardBars?: number | undefined | null): PatternMemoryResult
export interface PerformanceMetrics {
  /** Annualized Sharpe Ratio: (mean_return - risk_free) / std * sqrt(252) */
  sharpeRatio: number
  /** Annualized Sortino Ratio: (mean_return - risk_free) / downside_std * sqrt(252) */
  sortinoRatio: number
  /** Calmar Ratio: annualized_return / max_drawdown */
  calmarRatio: number
  /** Maximum Drawdown (as positive fraction, e.g. 0.25 = 25%) */
  maxDrawdown: number
  /** Maximum Drawdown duration in bars */
  maxDrawdownDuration: number
  /** Total cumulative return (e.g. 0.50 = 50%) */
  totalReturn: number
  /** Annualized return */
  annualizedReturn: number
  /** Annualized volatility (std of returns * sqrt(252)) */
  annualizedVolatility: number
  /** Win rate: fraction of positive returns */
  winRate: number
  /** Profit factor: sum of gains / sum of losses */
  profitFactor: number
  /** Average win / average loss ratio */
  payoffRatio: number
  /** Number of trading periods */
  numPeriods: number
  /** Skewness of returns */
  skewness: number
  /** Excess kurtosis of returns */
  kurtosis: number
  /** Value at Risk (5th percentile of returns) */
  var95: number
  /** Conditional VaR / Expected Shortfall (mean of returns below VaR) */
  cvar95: number
}
/**
 * Compute comprehensive performance metrics from a returns series.
 *
 * Input: array of period returns (e.g. daily returns as decimals: 0.01 = 1%)
 *
 * Parameters:
 * - returns: array of period returns
 * - risk_free_rate: annualized risk-free rate (default: 0.02 = 2%)
 * - periods_per_year: trading periods per year (default: 252 for daily)
 */
export declare function performanceMetrics(returns: Array<number>, riskFreeRate?: number | undefined | null, periodsPerYear?: number | undefined | null): PerformanceMetrics
/** Quick Sharpe Ratio calculation from returns. */
export declare function sharpeRatio(returns: Array<number>, riskFreeRate?: number | undefined | null, periodsPerYear?: number | undefined | null): number
/** Quick Sortino Ratio calculation from returns. */
export declare function sortinoRatio(returns: Array<number>, riskFreeRate?: number | undefined | null, periodsPerYear?: number | undefined | null): number
/** Quick Max Drawdown calculation from returns. */
export declare function maxDrawdown(returns: Array<number>): number
export interface PortfolioStats {
  /** Expected daily return of the portfolio */
  expectedReturnDaily: number
  /** Expected annualized return */
  expectedReturnAnnual: number
  /** Daily portfolio volatility (std dev) */
  volatilityDaily: number
  /** Annualized portfolio volatility */
  volatilityAnnual: number
  /** Daily portfolio variance */
  varianceDaily: number
  /** Sharpe ratio (annualized) */
  sharpeRatio: number
}
export interface CovarianceResult {
  /** Covariance matrix (flat, row-major, n_assets x n_assets) */
  covariance: Array<number>
  /** Correlation matrix (flat, row-major, n_assets x n_assets) */
  correlation: Array<number>
  /** Mean daily return per asset */
  meanReturns: Array<number>
  /** Annualized volatility per asset */
  volatilities: Array<number>
  /** Number of assets */
  nAssets: number
}
export interface EfficientFrontierPoint {
  /** Target return (annualized) */
  targetReturn: number
  /** Portfolio volatility at this point (annualized) */
  volatility: number
  /** Optimal weights for this point */
  weights: Array<number>
  /** Sharpe ratio at this point */
  sharpeRatio: number
}
export interface EfficientFrontierResult {
  /** Points along the efficient frontier */
  frontier: Array<EfficientFrontierPoint>
  /** Global Minimum Variance Portfolio */
  gmvp: EfficientFrontierPoint
  /** Maximum Sharpe Ratio (tangency) portfolio */
  maxSharpe: EfficientFrontierPoint
}
/**
 * Compute covariance matrix, correlation matrix, and per-asset stats
 * from multiple return series.
 *
 * Input: flat array of returns, row-major, with n_assets per row.
 * Each row = one time period, columns = assets.
 * Example: [ret_a_0, ret_b_0, ret_c_0, ret_a_1, ret_b_1, ret_c_1, ...]
 */
export declare function covarianceMatrix(returnsFlat: Array<number>, nAssets: number): CovarianceResult
/**
 * Compute portfolio return and risk for given weights.
 *
 * - returns_flat: flat return series (row-major, n_assets per row)
 * - n_assets: number of assets
 * - weights: portfolio weights (must sum to ~1)
 * - risk_free_rate: annualized (default: 0.02)
 */
export declare function portfolioStats(returnsFlat: Array<number>, nAssets: number, weights: Array<number>, riskFreeRate?: number | undefined | null): PortfolioStats
/**
 * Compute the efficient frontier using the analytical Markowitz solution.
 *
 * - returns_flat: flat return series (row-major, n_assets per row)
 * - n_assets: number of assets
 * - n_points: number of points on the frontier (default: 50)
 * - risk_free_rate: annualized (default: 0.02)
 */
export declare function efficientFrontier(returnsFlat: Array<number>, nAssets: number, nPoints?: number | undefined | null, riskFreeRate?: number | undefined | null): EfficientFrontierResult
export declare function pivotPoints(data: Array<MarketData>): Array<number>
export interface RegimeLeverageResult {
  /** Hybrid oscillator values (smoothed) */
  oscillator: Array<number>
  /** Yang-Zhang volatility (annualized) */
  yzVolatility: Array<number>
  /** Volatility percentile (0-1, rolling 252-bar) */
  volPercentile: Array<number>
  /** Regime label: 0=Defensive, 1=Moderate, 2=Bullish, 3=Aggressive */
  regime: Array<number>
  /** Leverage factor: 0.0, 1.0, 2.0, or 3.0 */
  leverage: Array<number>
  /** VIX ratio (vix/vix3m) if VIX data provided, else NaN */
  vixRatio: Array<number>
}
/**
 * Market Regime Adaptive Leverage System (MRALS)
 *
 * Classifies market into 4 regimes and assigns leverage:
 * - Aggressive (3x): low vol + bullish trend + normal VIX structure
 * - Bullish (2x): positive trend, moderate volatility
 * - Moderate (1x): neutral conditions
 * - Defensive (0x): high volatility or bearish signals
 *
 * Uses a hybrid oscillator combining:
 * - Price momentum (EMA fast/slow differential)
 * - Relative strength (21-bar return vs rolling mean, z-scored)
 * - Volatility component (VIX ratio deviation, z-scored if available)
 *
 * Parameters:
 * - data: OHLCV market data
 * - vix_values: optional VIX index values (same length)
 * - vix3m_values: optional VIX3M index values (same length)
 * - yz_window: Yang-Zhang vol window (default: 21)
 * - ema_fast: fast EMA for oscillator (default: 8)
 * - ema_slow: slow EMA for oscillator (default: 21)
 * - oscillator_smooth: EMA smoothing for oscillator (default: 5)
 * - vol_lookback: rolling window for vol percentile (default: 252)
 * - trend_period: SMA period for price trend (default: 50)
 */
export declare function regimeLeverage(data: Array<MarketData>, vixValues?: Array<number> | undefined | null, vix3MValues?: Array<number> | undefined | null, yzWindow?: number | undefined | null, emaFast?: number | undefined | null, emaSlow?: number | undefined | null, oscillatorSmooth?: number | undefined | null, volLookback?: number | undefined | null, trendPeriod?: number | undefined | null): RegimeLeverageResult
export declare function relativeStrengthIndex(prices: Array<number>, period: number): Array<number>
export interface RviResult {
  /** RVI line values */
  rvi: Array<number>
  /** Signal line (4-period weighted moving average of RVI) */
  signal: Array<number>
}
/**
 * Relative Vigor Index (RVI)
 *
 * Measures the conviction of a price move by comparing the close-open range
 * to the high-low range. The idea: in uptrends, closes tend to be above opens,
 * and the opposite in downtrends.
 *
 * RVI = SMA(N, numerator) / SMA(N, denominator)
 * where:
 *   numerator = (Close - Open) + 2*(Close[-1] - Open[-1]) + 2*(Close[-2] - Open[-2]) + (Close[-3] - Open[-3]) / 6
 *   denominator = (High - Low) + 2*(High[-1] - Low[-1]) + 2*(High[-2] - Low[-2]) + (High[-3] - Low[-3]) / 6
 *
 * Signal = (RVI + 2*RVI[-1] + 2*RVI[-2] + RVI[-3]) / 6
 *
 * Parameters:
 * - data: OHLCV market data
 * - period: SMA smoothing period (default: 10)
 */
export declare function relativeVigorIndex(data: Array<MarketData>, period?: number | undefined | null): RviResult
export declare function simpleMovingAverage(data: Array<number>, period: number): Array<number>
export interface SpreadEstimatorResult {
  /** Rolling bid-ask spread estimates (0.01 = 1% spread) */
  spreads: Array<number>
  /** Rolling bid-ask spread with sign preserved */
  signedSpreads: Array<number>
}
/**
 * Rolling Bid-Ask Spread Estimator (Ardia, Guidotti & Kroencke, 2024)
 *
 * Estimates bid-ask spread from OHLC prices using moment conditions
 * and a rolling window. More accurate than Roll (1984), Corwin-Schultz (2012),
 * and Abdi-Ranaldo (2017) estimators, especially in low-liquidity markets.
 *
 * A returned value of 0.01 means a 1% spread.
 */
export declare function spreadEstimator(data: Array<MarketData>, window: number): SpreadEstimatorResult
/**
 * Classic Roll (1984) spread estimator for comparison.
 * spread = 2 * sqrt(-Cov(ΔP_t, ΔP_{t-1})) if covariance is negative, else 0.
 */
export declare function rollSpreadEstimator(prices: Array<number>, window: number): Array<number>
/**
 * Corwin-Schultz (2012) High-Low spread estimator.
 * Uses high and low prices over two consecutive periods.
 */
export declare function corwinSchultzSpreadEstimator(data: Array<MarketData>, window: number): Array<number>
export declare function stochasticMomentumIndex(data: Array<MarketData>, lookbackPeriod?: number | undefined | null, firstSmoothing?: number | undefined | null, secondSmoothing?: number | undefined | null): Array<number>
export declare function stochasticOscillator(data: Array<MarketData>, period: number): Array<number>
export interface ThreeWayResult {
  /** Combined score at each bar (-3 to +3) */
  score: Array<number>
  /** Trend component: +1 (SMA fast > slow), -1 (opposite), 0 (warmup) */
  trend: Array<number>
  /** Momentum component: +1 (RSI > 50), -1 (RSI < 50), 0 (neutral/warmup) */
  momentum: Array<number>
  /** Volatility component: +1 (expanding, ATR rising), -1 (contracting), 0 (warmup) */
  volatility: Array<number>
  /** Signal: 1 = strong buy (score >= 2), -1 = strong sell (score <= -2), 0 = neutral */
  signals: Array<number>
}
/**
 * Three Way Indicator
 *
 * Combines three independent market dimensions into a single composite score:
 * 1. Trend: SMA crossover (fast vs slow)
 * 2. Momentum: RSI position relative to 50
 * 3. Volatility: ATR direction (expanding or contracting)
 *
 * Score ranges from -3 (all bearish) to +3 (all bullish).
 * Signals fire when score >= buy_threshold or <= -sell_threshold.
 *
 * Parameters:
 * - data: OHLCV market data
 * - fast_sma: fast SMA period (default: 10)
 * - slow_sma: slow SMA period (default: 30)
 * - rsi_period: RSI period (default: 14)
 * - atr_period: ATR period (default: 14)
 * - atr_lookback: bars to compare ATR direction (default: 5)
 * - signal_threshold: absolute score threshold for signals (default: 2)
 */
export declare function threeWayIndicator(data: Array<MarketData>, fastSma?: number | undefined | null, slowSma?: number | undefined | null, rsiPeriod?: number | undefined | null, atrPeriod?: number | undefined | null, atrLookback?: number | undefined | null, signalThreshold?: number | undefined | null): ThreeWayResult
export declare function trendsMeter(data: Array<MarketData>, period?: number | undefined | null): Array<number>
export interface VolatilityBucket {
  /** "low", "medium", or "high" */
  regime: string
  /** ATR multiplier for this regime */
  atrMultiplier: number
  /** Current ATR value */
  atr: number
  /** Current volatility (rolling std of returns) */
  volatility: number
  /** Stop-loss distance = ATR * multiplier */
  stopDistance: number
  /** Low volatility threshold (percentile) */
  lowThreshold: number
  /** High volatility threshold (percentile) */
  highThreshold: number
}
export interface VolatilityEngineResult {
  /** ATR values (full length, NaN for warmup) */
  atr: Array<number>
  /** Rolling volatility (std dev of returns, full length, NaN for warmup) */
  volatility: Array<number>
  /** Volatility regime at each bar: 0=low, 1=medium, 2=high, -1=warmup */
  regimes: Array<number>
  /** ATR multiplier selected at each bar */
  atrMultipliers: Array<number>
  /** Stop-loss distance at each bar (ATR * multiplier) */
  stopDistances: Array<number>
  /** Percentile-based low threshold at each bar */
  lowThresholds: Array<number>
  /** Percentile-based high threshold at each bar */
  highThresholds: Array<number>
}
/**
 * Volatility-Adaptive Engine
 *
 * Computes ATR + rolling volatility (std dev of returns), then classifies
 * each bar into a volatility regime (low/medium/high) using rolling percentiles.
 * Each regime maps to a different ATR multiplier for dynamic stop-loss sizing.
 *
 * Parameters:
 * - data: OHLCV market data
 * - atr_period: ATR lookback (default: 14)
 * - vol_period: rolling std dev period for returns (default: 20)
 * - vol_history_len: number of bars for percentile calculation (default: 200)
 * - vol_warmup: minimum history before assigning regimes (default: 50)
 * - percentile_low: low vol threshold percentile (default: 20)
 * - percentile_high: high vol threshold percentile (default: 80)
 * - low_vol_mult: ATR multiplier for low volatility (default: 1.5)
 * - med_vol_mult: ATR multiplier for medium volatility (default: 2.5)
 * - high_vol_mult: ATR multiplier for high volatility (default: 4.0)
 */
export declare function volatilityEngine(data: Array<MarketData>, atrPeriod?: number | undefined | null, volPeriod?: number | undefined | null, volHistoryLen?: number | undefined | null, volWarmup?: number | undefined | null, percentileLow?: number | undefined | null, percentileHigh?: number | undefined | null, lowVolMult?: number | undefined | null, medVolMult?: number | undefined | null, highVolMult?: number | undefined | null): VolatilityEngineResult
/** Get a single volatility bucket classification for the current bar */
export declare function volatilityBucket(currentAtr: number, currentVolatility: number, volatilityHistory: Array<number>, percentileLow?: number | undefined | null, percentileHigh?: number | undefined | null, lowVolMult?: number | undefined | null, medVolMult?: number | undefined | null, highVolMult?: number | undefined | null): VolatilityBucket
export interface YangZhangResult {
  /** Yang-Zhang volatility (annualized) */
  volatility: Array<number>
  /** Overnight component (close-to-open) */
  overnightVol: Array<number>
  /** Intraday component (open-to-close) */
  intradayVol: Array<number>
  /** Rogers-Satchell component */
  rogersSatchell: Array<number>
}
/**
 * Yang-Zhang Volatility Estimator
 *
 * Combines three volatility components for more accurate estimation than
 * simple standard deviation:
 * - Overnight volatility: log(Open / prev_Close)²
 * - Intraday volatility: log(Close / Open)²
 * - Rogers-Satchell: log(H/O)*log(H/C) + log(L/O)*log(L/C)
 *
 * YZ = sqrt(overnight + k * intraday + (1-k) * RS)
 * where k = 0.34 / (1.34 + (n+1)/(n-1))
 *
 * Output is annualized (multiplied by sqrt(252)).
 *
 * Parameters:
 * - data: OHLCV market data
 * - window: rolling window (default: 10)
 */
export declare function yangZhangVolatility(data: Array<MarketData>, window?: number | undefined | null): YangZhangResult
export declare function sum(a: number, b: number): number