[Scilab-users] confidence interval for lsqrsolve solution?
Ingo
ingo at drsteiner.de
Fri Apr 25 11:07:11 CEST 2014
Hello everyone,
I have done some data analysis using lsqrsolve(), which has delivered very
good (small value of optimization criteria), stable (same solution from
different initial parameters) and fast (small number of iterations) results.
However, I could not find out how to retrieve confidence bands or error
estimates of the final parameter set. Is this possible? Can I access the
gradient matrix of the last iteration and compute the confidence bands from
there?
Thanks for any hint,
Ingo
--
View this message in context: http://mailinglists.scilab.org/confidence-interval-for-lsqrsolve-solution-tp4030375.html
Sent from the Scilab users - Mailing Lists Archives mailing list archive at Nabble.com.
More information about the users
mailing list