[ncl-talk] function "bootstrap_regcoef"

[Dennis Shea]

I see that the function documentation

	function bootstrap_regcoef (
		*x*         : numeric,    <== parameter 0
		y         : numeric,    <== parameter 1
		nBoot [1] : integer,
		nDim  [*] : integer,
		opt   [1] : logical
	)

allows an 'x' with multiple dimensions.

This is **NOT** correct. (My fault!)

If you look at:

%> less $NCARG_ROOT/lib/ncarg/nclscripts/csm/bootstrap.ncl

The actual function is prototyped as one dimensional [*]

function bootstrap_regcoef(*x[*]*:numeric, z:numeric,
nBoot[1]:integer, nDim:integer, opt[1]:logical)

=============================================

The bootstrap_regcoef examples at

   http://www.ncl.ucar.edu/Applications/bootstrap.shtml

all have the first (0-th) argument as one dimensional.

==============================================

The documentation has been changed:
    http://www.ncl.ucar.edu/Document/Functions/Bootstrap/bootstrap_regcoef.shtml

====

I note you refer to 'time' as  'year*month' In virtually all NCL
function 'time' is one dimensional.

THX









On Mon, Dec 11, 2017 at 11:36 PM, 王健 <montager.wang at gmail.com> wrote:

> Hi there,
>
> When I use function "bootstrap_regcoef" to estimate regression
> coefficients of Y(year*month*lat*lon) on X (year*month*lat*lon),
> here's the error info:
> "fatal:Number of dimensions in parameter (0) of (bootstrap_regcoef) is
> (2), (1) dimensions were expected"
>
> while I use "bootstrap_correl" on the same data, everything seems okay,
> is there a bug or something in " bootstrap_regcoef" function?
>
> Looking forward to your response!
>
> Best regard,
>
> Chris
>
> _______________________________________________
> 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/20171212/d8cd7494/attachment.html>