Re: Statistical significance

Hi Guilherme,
I do not see anything wrong with your script, other than this line not
being necessary:
plot(k-1) = plot(k-1)

Significance contours can get rather dense and convoluted especially for
high resolution grids, but the plotting should not take anywhere close
to 4 hours.

I would suggest trying the following:
Change this:
res2@cnLevelSelectionMode = "ManualLevels"
res2@cnMinLevelValF = 0.00
res2@cnMaxLevelValF = 1.05
res2@cnLevelSpacingF = 0.01
to this:
res2@cnLevelSelectionMode = "ExplicitLevels"
res2@cnLevels = niv_sig

As you're only interested in shading all areas less than niv_sig there's
no reason to specify 106 contour levels. I am wondering if NCL still
calculates their placement even though you're not drawing them. If the
above changes do not speed things up, please respond to NCL-talk and let
everyone know..
Adam

On 12/01/2013 06:09 PM, Guilherme Martins wrote:
> Hi all,
>
> I'm calculating the correlation, but when I inserted the statistical
> significance using the rtest, my code became very slow, without the
> statistical significance everything is done during one minute. With
> the statistical significance, the time is of four hours. I don't know
> where I'm doing something wrong.
>
> The lines associated with the statistical significance are: 143-146.
>
> plot1 = gsn_csm_contour(wks,m1,res2)
> plot1 = ShadeLtContour(plot1,niv_sig,17)
> overlay (plot(k-1), plot1)
> plot(k-1) = plot(k-1)
>
> *My script is in attached.*
>
> Thanks,
>
> Guilherme.
>
> --
> *Guilherme Martins*
> https://sites.google.com/site/jgmsantos/
>
>
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

-- 
______________________________________________________________
Adam Phillips                                asphilli@ucar.edu
NCAR/Climate and Global Dynamics Division       (303) 497-1726
P.O. Box 3000				
Boulder, CO 80307-3000    http://www.cgd.ucar.edu/cas/asphilli

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Dec 2 13:24:39 2013