calling method
(1)C=xcov(A, b): when a and b are vectors with length m (m > 1), the returned result is the covariance function column vector with length 2M-1;
(2)C=xcov(A): estimate the autocorrelation function sequence of vector a;
(3)C=xcov(A): when a is a matrix of M×N, a matrix with (2M-l) rows and N2 columns is returned, and the columns of the matrix are composed of cross-covariance functions among all columns of the matrix A;
(4)C=xcov (,maxlag): returns the covariance function sequence with the length of 2*maxlag+1, from -maxlag to max-lag, and the default value of maxlag is m-1;
(5) [c, lags] = xcov (…): the coordinate vector lags is also returned.
(6)C=xcov(…, scaleopt): the parameter scaleopt is used to specify the estimation method for covariance function estimation. That is,
biased: biased estimation mode;
unbiased: unbiased estimation method;
coeff normalizes the sequence to ensure that the autocorrelation sequence of the sample value with zero lag is always 1;
none: calculate the non-normalized correlation of the sequence, which is the default way.