[ncl-talk] Missing Regression Coefficient Confidence Intervals in reg_multlin_stats function

[Dennis Shea]

Currently, there are no plans to alter the underlying f77 subroutine used
by NCL's
   https://www.ncl.ucar.edu/Document/Functions/
Contributed/reg_multlin_stats.shtml

This function returns (as attributes) *a lot of information including and
ANOVA table*.

Both examples associated with the function's documentation are compared to
R results. The 2nd example's R results  provide confidence results.

I would assume that the ANOVA information returned by the function could be
used to derive confidence limits associated with each 'b'.   Note: I can
not tell you how.

Good luck




On Mon, Oct 16, 2017 at 9:19 AM, Andreas Chrysanthou <eeac at leeds.ac.uk>
wrote:

> Hi NCL users,
>
> I’m currently using the function reg_multlin_stats to perform multiple
> linear regression analysis with a number of independent variables.
>
> Unlike the redline_stats function that in NCL v6.4.0 which has the
> regression coefficient confidence intervals  the multiple linear regression
> do not produce those as output.
>
> Is there any plan to add those statistics in a next version of NCL?
>
> Any suggestions?
>
> Cheers,
> Andreas
>
>
>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20171016/bbad7b0a/attachment.html>