2nd try : Nearest defined neighbor function

[Patrick Reuter]

Dear Stephen and Pablo,

that is exactly what I was looking for, thank you. I discovered defval
and it does the job. I assume that now I have to loop through the XY
grid values, and "defval" the undefined values manually for each grid
point -- until somebody knows how to make a UDF out of it.

Happy new year
    Patrick


"Stephen R McMillan" <smcmillan at PLANALYTICS.COM> a écrit :

> Patrick,
> What if an undefined point has more than one "nearest neighbor" and are of
> different values?  Do you want to assign an average of these, the highest,
> lowest, etc.?  I'm assuming each gridpoint has up to eight nearest
> neighbors in complex coastal areas.  For example, you could get the max
> nearest neighbor using something like this:
>
> 'define newvar= max(max(var,x=-152,x=-150),y=262,y=264)'
>
> Then assign result to point (in this case, x=-151, y=263) using the 'set
> defval var...' command.
>
> Similarly, you could use 'define var= min(min...' to get min value of
> nearest 8-neighbor or the "aave" command to get average of nearest
> neighbor.
>
> Hope this helps....
> Stephen McMillan
>
>
>
>
> Patrick Reuter <preuter at LABRI.FR>
> Sent by: GRADSUSR at LIST.CINECA.IT
> 01/06/2009 07:39 AM
> Please respond to
> GRADSUSR at LIST.CINECA.IT
>
>
> To
> GRADSUSR at LIST.CINECA.IT
> cc
>
> Subject
> 2nd try : Nearest defined neighbor function
>
>
>
>
>
> Sorry for multiples, message was somehow rejected ..
>
> Dear all,
>
> I have a variable which is only defined over the sea, elsewhere there
> are undefined values. Is there a function in Grads that can draw me a
> map where at locations of undefined values, instead of leaving it
> blank, draws me the value of the nearest defined value of the
> 8-neighborhood ? I know about the existence of 'smth9', but it does
> not exactly what I want, since it smoothes out, and 'const', since you
> have to specify a default value which I don't know since it should to
> be determined from the nearest neighbor !
>
> Can anybody help?
>
> Thanks in advance
>
> Patrick
>
>
>
>
> ***************************************************
> The information contained in this e-mail message is intended only
> for the use of the recipient(s) named above and may contain
> information that is privileged, confidential, and/or proprietary.
> If you are not the intended recipient, you may not review, copy or
> distribute this message.  If you have received this communication in
> error, please notify the sender immediately by e-mail, and delete
> the original message.
> ***************************************************