Re: [las_users] How does subsetting curvilinear plots work in LAS?

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi Anton,
The method is quite crude and does not always do a good job. It uses the 
method in this Ferret FAQ,
http://ferret.pmel.noaa.gov/Ferret/faq/plotting-subsets-of-data-on-a-curvilinear-grid, 
using a coarse grid and calling our regridding function just to try to 
determine the range of index values to use in making a subset, still in 
the curvilinear coordinate system. Then we load that I-J subset of the 
coordinate variables and the data, still making the plot in curvilinear 
coordinates. For global grids such as the tripolar grid, the method 
works better than it does for some of the localized coastal grids.

In PyFerret, we are experimenting with new regridding functions that are 
quite fast and which should greatly improve this task.

-Ansley


On 5/20/2013 8:07 PM, Anton Yoel Buenavista wrote:
> Dear LAS mailing list,
>
> I'm wondering how subsetting curvilinear plots in LAS works. 
> Particularly, if the LAS selection box exceeds the bounds of the 
> regridded curvilinear domain, what formula does Ferret use to 
> transform the "out of bounds" coordinates to the proper values for 
> Ferret's internal use?
>
> Attached are some screenshots to show an example. On [screen_1.png], 
> I've done a rubberband selection on the mini-map with the lower-left 
> portion being outside the dataset domain. On [screen_2.png], you can 
> clearly see that the "out of bounds" area has been properly adjusted 
> to be within the domain. Kindly take note that this is a curvilinear 
> dataset.
>
> In addition, if I make a rubberband selection, does Ferret regrid the 
> entire domain first and then subset from there, or does LAS pass the 
> selection lon/lat coordinates to Ferret and Ferret transforms the 
> coordinates to the x/y equivalent in order to subset the curvilinear 
> dataset before regridding? Any information is greatly appreciated.
> --
> Regards,
> Anton Buenavista