@resample |

Syntax: @resample(m1, n2, n3, v4)

Input 1: matrix m1

Input 2: (optional) integer n2

Input 3: (optional) positive integer n3

Input 4: (optional) vector v4

Output: matrix

This function returns a matrix whose rows are randomly drawn with replacement from rows of the input matrix.

represents the number of “extra” rows to be drawn from the matrix. If the input matrix has r rows and c columns, the output matrix will have rows and columns. By default, .

represents the block size for the resample procedure. If you specify , then blocks of consecutive rows of length will be drawn with replacement from the first rows of the input matrix.

You may provide a name for the vector to be used for weighted resampling. The weighting vector must have length and all elements must be non-missing and non-negative. If you provide a weighting vector, each row of the input matrix will be drawn with probability proportional to the weights in the corresponding row of the weighting vector. (The weights need not sum to 1. EViews will automatically normalize the weights).

matrix xb = @bootstrap(x)

To draw without replacement from rows of a matrix, use @permute.