/**
 * Variance: compute a single variance over the whole series, dropping NaNs.
 * Uses `ddof` (delta degrees of freedom) like NumPy/pandas: denominator = N - ddof.
 * Default pandas behavior is sample variance (ddof=1), so the default here is ddof=1.
 */
export declare function variance(source: ArrayLike<number>, options?: {
    ddof?: number;
    skipna?: boolean;
}): number;
/**
 * Standard deviation: compute a single standard deviation over the whole series, dropping NaNs.
 * Uses `ddof` (delta degrees of freedom) like NumPy/pandas: denominator = N - ddof.
 * Default pandas behavior is sample standard deviation (ddof=1), so the default here is ddof=1.
 */
export declare function stdev(source: ArrayLike<number>, options?: {
    ddof?: number;
    skipna?: boolean;
}): number;
/**
 * Pairwise covariance: cov(X,Y) = E[XY] - E[X]E[Y]
 */
export declare function covar(x: ArrayLike<number>, y: ArrayLike<number>, options?: {
    ddof?: number;
}): number;
/**
 * Rolling variance over a sliding window. Accepts `{ skipna?, ddof? }`.
 * @param source Input array
 * @param period Window length (>0)
 * @param options Optional `{ skipna?, ddof? }`
 * @returns Float64Array of rolling variances
 */
export declare function rollvar(source: ArrayLike<number>, period: number, options?: {
    skipna?: boolean;
    ddof?: number;
}): Float64Array;
export declare function rollcovar(x: ArrayLike<number>, y: ArrayLike<number>, period: number, options?: {
    ddof?: number;
}): Float64Array;
export declare function rollstdev(source: ArrayLike<number>, period: number, options?: {
    skipna?: boolean;
    ddof?: number;
}): Float64Array;