co_matrices¶
-
pyoof.
co_matrices
(res, jac, n_pars)[source] [edit on github]¶ Computes the estimated Variance-Covariance and the Correlation matrices. Assuming an estimator normally distributed (and consistent).
Parameters: res :
ndarray
Last residual evaluation from a fit procedure (least squares minimization), residual understood as data - model.
jac :
ndarray
Last Jacobian matrix evaluation from a fit procedure.
n_pars :
int
Total number of parameters in the model (only the ones that have been fitted). It is related to the degrees of freedom.
Returns: cov :
ndarray
Variance-Covariance matrix. Dimensions \(n \times p\).
corr :
ndarray
Correlation matrix. Dimensions \(n \times p\).