Language Reference

COVLAG Function

computes autocovariance estimates for a vector time series

COVLAG( x, k)

The inputs to the COVLAG function are as follows:


x
is an n x nv matrix of time series values; n is the number of observations, and nv is the dimension of the random vector.

k
is a scalar, the absolute value of which specifies the number of lags desired. If k is positive, a mean correction is made. If k is negative, no mean correction is made.
The COVLAG function computes a sequence of lagged crossproduct matrices. This function is useful for computing sample autocovariance sequences for scalar or vector time series.

The value returned by the COVLAG function is an nv x (k*nv) matrix. The ith nv x nv block of the matrix is the sum
\frac{1}n \sum_{j=i}^n x_j^' x_{j-i+1}  {if } k\lt
where x_j is the jth row of x. If k > 0, then the ith nv x nv block of the matrix is
\frac{1}n \sum_{j=i}^n (x_j-\bar{x})^'(x_{j-i+1}-\bar{x})
where \bar{x} is a row vector of the column means of x. For example, the following statements produce the matrix {cov}, as shown:
  
    x={-9,-7,-5,-3,-1,1,3,5,7,9}; 
    cov=covlag(x,4);
 

  
               COV           1 row       4 cols    (numeric) 
  
                         33      23.1      13.6       4.9
 

Previous Page | Next Page | Top of Page