Usage: |
res = VaRfitcopula(history,copula,method)
|
Input: |
| history | n x 2 matrix, the data
|
| copula | scalar, integer between 1 and 22, selects the copula;
see VaRcopula for details
|
| method | scalar, selects the method for estimation
method=1 returns the least squares fit for the cdf
method=2 returns a copula maximum likelihood fit for the pdf,
using the Newton Method, fall back to 3 if it fails
method=3 returns a copula maximum likelihood fit for the pdf,
using an interval search, slower than 2, but more stable
method=4 returns a true maximum likelihood fit for the pdf
|
Output: |
| res | list containing the estimates of the copula parameter
and standard deviations of two marginal Normal distributions |