Re: problems with gsnAddCyclic

Hi Kyle,

Given that you have 1D lat/lon coordinate arrays, it seems odd that gsnAddCyclic would be causing this much problem.

You do *not* want to set gsnAddCyclic to True for regional data. I assume by "regional" in your email below, you meant that you just zoomed in on the map using mp resources?

Having access to your data and a script would be great, thanks, so we can take a look here.

--Mary

On Oct 29, 2012, at 8:31 AM, Kyle Griffin wrote:

> Hi all,
>
> In attempting to use gsnAddCyclic = True in a script, I have discovered that it does not seem to function properly with the data as I am trying to plot it. I've simplified down several portions of the script and it does appear to boil down to this resource.
>
> The data I'm plotting has printVarSummary statements that I've linked below represent each of the variables being plotted in the script (I've turned a couple off for simplicity's sake here as well). While the addCyclic documentation suggests it should be "True" by default anytime in can detect a cyclic dataset being plotted, this does not appear to be the case. Further, when it is set to true, it distorts the data and plots it at incorrect coordinates - earlier tests showed a full set of longitudes would be plotted without distortion, but shifted ~90 degrees to the west. I've linked to a pair of smaller, regional maps that demonstrate the issue at hand - one plotted correctly but without a cyclic point, and one with cyclic points turned on where data is distorted.
>
> If further tinkering with my script is necessary, I can upload the script (easy to set up for a different machine) and data file that could be used to test - just let me know. Finally, I've tested this on the 6.1.0-beta and the just-uploaded full 6.1.0 with the same effects. Any advice? Thanks...
>
> Images:
> http://www.atmos.albany.edu/student/kgriffin/tmp/jet_mslp_irrwind_wpac_568272.png
> http://www.atmos.albany.edu/student/kgriffin/tmp/jet_mslp_irrwind_wpac_568272_nocyclic.png
>
>
> Kyle
>
> p.s. Variables below are lonFlipped before plotting and thus the cyclic point is not being added at 180 in linked maps, despite being lon=0..359.5 below. Same errors occur with/without lonFlip.
>
> printVarSummary statements:
>
> Variable: jet
> Type: float
> Total Size: 1039680 bytes
> 259920 values
> Number of Dimensions: 3
> Dimensions and sizes: [time | 1] x [lat | 361] x [lon | 720]
> Coordinates:
> time: [77733.75..77733.75]
> lat: [-90..90]
> lon: [ 0..359.5]
> Number Of Attributes: 1
> _FillValue : 9.96921e+36
>
> Variable: pw
> Type: float
> Total Size: 1039680 bytes
> 259920 values
> Number of Dimensions: 3
> Dimensions and sizes: [time | 1] x [lat | 361] x [lon | 720]
> Coordinates:
> time: [77733.75..77733.75]
> lat: [-90..90]
> lon: [ 0..359.5]
> Number Of Attributes: 1
> _FillValue : 9.96921e+36
>
> Variable: pmsl
> Type: float
> Total Size: 1039680 bytes
> 259920 values
> Number of Dimensions: 3
> Dimensions and sizes: [time | 1] x [lat | 361] x [lon | 720]
> Coordinates:
> time: [77733.75..77733.75]
> lat: [-90..90]
> lon: [ 0..359.5]
> Number Of Attributes: 1
> _FillValue : 9.96921e+36
>
> Variable: irr
> Type: float
> Total Size: 2079360 bytes
> 519840 values
> Number of Dimensions: 4
> Dimensions and sizes: [component | 2] x [time | 1] x [lat | 361] x [lon | 720]
> Coordinates:
> component: [0..1]
> time: [77733.75..77733.75]
> lat: [-90..90]
> lon: [ 0..359.5]
> Number Of Attributes: 4
> units : m/s
> long_name : divergent zonal [0] and meridional [1] winds
> average_op_ncl : dim_avg_n over dimension(s): pres_lev
> _FillValue : 9.96921e+36
> ----------------------------------------
> Kyle S. Griffin
> Department of Atmospheric and Oceanic Sciences
> University of Wisconsin - Madison
> 1225 W Dayton St, Madison, WI 53706
> Room 1421 Email: ksgriffin2@wisc.edu
>
> _______________________________________________
> 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 Mon Oct 29 09:29:01 2012