Re: help with RMS circles in Taylor plot example

From: Dennis Shea <shea_at_nyahnyahspammersnyahnyah>
Date: Wed Oct 02 2013 - 07:40:50 MDT

THX ... nice plots.

I have not read the Taylor paper in a long time so I do not remember the
details.

The pattern correlations are centered. I would think the RMSE circles
are centered.

D

On 10/1/13 1:59 PM, Noble, Erik U. (GISS)[COLUMBIA UNIVERSITY] wrote:
> Dear Dennis,
> I was able to make the plots using the RMSE code. The picture is attached
> and I will send my code so that people can use.
>
> But, Taylor 2001 discusses that the RMSE circles used in their Taylor
> diagrams are centered RMSE, RMSE with the mean bias removed.
>
> The Taylor diagram plot allows us to 'hard-code' the RMSE semicircles, and
> that is what I am showing here. These RMSE cycles are not the centered
> RMSE, right?
> Sincerely,
> Erik
>
>
>
> On 9/18/13 5:27 PM, "Dennis Shea" <shea@ucar.edu> wrote:
>
>> Hi Erik
>>
>> Unfortunately, the concentric circles were written for an X-axis
>> of max 1.65: specifically: xyMax = 1.65
>>
>> Since your plot extends to 2.0 you will have to copy the
>> downloadable taylor code and change section code: Part 3
>>
>> ; ----------------------------------------------------------------
>> ; Part 3:
>> ; Concentric about 1.0 on XB axis
>> ; I think this is correct. Still test mode.
>> ; ----------------------------------------------------------------
>> [snip]
>> dum9 = new(ncon,graphic)
>>
>> do n=1,ncon
>> rr = n*dx ; radius from 1.0 [OBS] abscissa
>> xx = 1. + rr*cos(ang)
>> yy = fabs( rr*sin(ang) )
>> if (n.le.2) then
>> dum9(n-1) = gsn_add_polyline(wks,taylor,xx,yy,respl)
>> end if
>> if (n.eq.3) then
>> n3 = floattointeger( 0.77*npts )
>> dum9(n-1) =
>> gsn_add_polyline(wks,taylor,xx(0:n3),yy(0:n3),respl)
>> end if
>> if (n.eq.4) then
>> n4 = floattointeger( 0.61*npts )
>> dum9(n-1) =
>> gsn_add_polyline(wks,taylor,xx(0:n4),yy(0:n4),respl)
>> end if
>> end do
>> [snip]
>>
>> You will have to manually change n3 ans n4 by altering the 0.77 and 0.61
>> factors. I am sure that there is some objective way to terminate but I
>> don't know what it might be.
>>
>> If you get it working .. send in the modes. Somebody else
>> might be ables to use it.
>>
>> Regards
>> D
>>
>>
>>
>>
>>
>> On 9/18/13 2:58 PM, Noble, Erik U. (GISS)[COLUMBIA UNIVERSITY] wrote:
>>>
>>> Dear NCL
>>>
>>> Attached is an image I made from the NCL Taylor plot examples.
>>>
>>>
>>> Here is the part of the taylor_plot.ncl script that deals with the RMS
>>> circles.
>>>
>>>
>>> My only problem is that I cannot get the RMS lines to draw correctly on
>>> the plot. How do I prevent them from extending past the
>>> outer-correlation arch on the plot?
>>>
>>>
>>> Thank you.
>>>
>>> Sincerely,
>>>
>>> Erik
>>>
>>>
>>> ; ----------------------------------------------------------------
>>> ; Part 3:
>>> ; Concentric about 1.0 on XB axis
>>> ; I think this is correct. Still test mode.
>>> ; ----------------------------------------------------------------
>>> if (rOpts .and. isatt(rOpts,"centerDiffRMS") \
>>> .and. rOpts@centerDiffRMS) then
>>> respl = True ; polyline mods
>>> desired
>>> respl@xyLineThicknessF = 1.0 ; line thickness
>>> respl@xyLineDashPattern = 2 ; short dash lines
>>> respl@gsLineColor = "Black" ; line color
>>> if (isatt(rOpts,"centerDiffRMS_color")) then
>>> respl@gsLineColor = "LightGray"
>>> end if
>>>
>>> dx = 0.25
>>> ncon = 8 ; 0.75, 0.50,
>>> 0.25, 0.0
>>> npts = 100 ; arbitrary
>>> ang = fspan(180,360,npts)*rad
>>>
>>> dum9 = new(ncon,graphic)
>>> ;;;; IF YOU DESIRE RMS LABELS
>>> rms_labels = (/0.0,0.25,0.50,0.75,1.0,1.25,1.50,1.75,2.0/)
>>>
>>> do n=1,ncon
>>> respl@gsLineLabelString = rms_labels(n-1) ;IF YOU
>>> DESIRE RMS LABELS
>>> respl@gsLineLabelFontColor = "LightGray"
>>> rr = n*dx ; radius from 1.0 [OBS] abscissa
>>> xx = 1. + rr*cos(ang)
>>> yy = fabs( rr*sin(ang) )
>>> if (n.le.2) then
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>> if (n.eq.3) then
>>> n3 = floattointeger( 0.61*npts )
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx(0:n3),yy(0:n3),respl)
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>> if (n.eq.4) then
>>> n4 = floattointeger( 0.61*npts )
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx(0:n4),yy(0:n4),respl)
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>>
>>> if (n.eq.5) then
>>> n5 = floattointeger( 0.22*npts )
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx(0:n5),yy(0:n5),respl)
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>> if (n.eq.6) then
>>> n6 = floattointeger( 0.61*npts )
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx(0:n6),yy(0:n6),respl)
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>> if (n.eq.7) then
>>> n7 = floattointeger( 0.61*npts )
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx(0:n7),yy(0:n7),respl)
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>> if (n.le.8) then
>>> n8 = floattointeger( 0.71*npts )
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx(0:n8),yy(0:n8),respl)
>>> dum9(n-1) =
>>> gsn_add_polyline(wks,taylor,xx,yy,respl)
>>> end if
>>>
>>> end do
>>> delete(ang)
>>> delete(xx)
>>> delete(yy)
>>> taylor@$unique_string("dum")$ = dum9
>>>
>>> end if
>>>
>>> _______________________________________________________
>>> Erik U. Noble
>>>
>>> Tel: +1 (212) 678-5512
>>> Fax: +1 (212) 678-5552
>>> E-mail: erik.noble@nasa.gov
>>> WEB: http://www.giss.nasa.gov/staff/enoble.html
>>> GISS: http://www.giss.nasa.gov
>>> _______________________________________________________
>>>
>>> * Open in Google Docs Viewer
>>> * Open link in new tab
>>> * Open link in new window
>>> * Open link in new incognito window
>>> *
>>> * Download file
>>> * Copy link address
>>> * Edit PDF File on PDFescape.com
>>> * Open in Google Docs Viewer
>>> * Open link in new tab
>>> * Open link in new window
>>> * Open link in new incognito window
>>> *
>>> * Download file
>>> * Copy link address
>>> * Edit PDF File on PDFescape.com
>>>
>>>
>>>
>>> _______________________________________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Wed Oct 2 07:41:00 2013

This archive was generated by hypermail 2.1.8 : Fri Oct 04 2013 - 16:45:17 MDT