comp.lang.idl-pvwave archive: archive » Re: BYTSCL and NAN keyword

Home » Public Forums » archive » Re: BYTSCL and NAN keyword

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: BYTSCL and NAN keyword [message #38280 is a reply to message #38279]

Tue, 02 March 2004 09:18

David Fanning
Messages: 11724
Registered: August 2001

Senior Member

Kenneth Bowman writes:

> Has anyone else noticed this problem with BYTSCL and the NAN keyword?
>
> The BYTSCL function scales integers and floats into bytes between 0 and
> 255 (or the value set by the TOP keyword). If the NAN keyword is set,
> NANs in the input array are set to 0 in the output array. But 0 falls
> into the valid range for good values (0 to TOP)!
>
> Because it is not possible to set NANs to a value outside the valid
> range (greater than TOP), it is not possible to distinguish missing from
> valid data.
>
> As best I can tell, the only solution is to not use the NAN keyword and
> scale the valid data only by using WHERE to find all the valid values.
> Am I missing something obvious? (Quite possible, I admit.)

Well, NANs are definitely floats (that is, they have a float-type
bit pattern). So it seems reasonable to me that if you are
trying to stuff it into a byte (which can only have values between
0 and 255, that 0 is a good choice. (The only possible other
choice is 255, but that just turns your problem on its head.)

I think what I would do is locate the NANs and save their
indices. Then scale my data into 255 colors (0 to 254), leaving
color index 255 for the NAN color (whatever that is). This will
take a couple of steps, but that's what IDL programs are for. :-)

Cheers,

David

--
David Fanning, Ph.D.
Fanning Software Consulting
Coyote's Guide to IDL Programming: http://www.dfanning.com/

Report message to a moderator

[Message index]

		Re: BYTSCL and NAN keyword By: David Fanning on Tue, 02 March 2004 15:48
		Re: BYTSCL and NAN keyword By: Kenneth P. Bowman on Tue, 02 March 2004 15:18
		Re: BYTSCL and NAN keyword By: David Fanning on Tue, 02 March 2004 14:07
		Re: BYTSCL and NAN keyword By: JD Smith on Tue, 02 March 2004 13:32
		Re: BYTSCL and NAN keyword By: Craig Markwardt on Tue, 02 March 2004 09:39
		Re: BYTSCL and NAN keyword By: David Fanning on Tue, 02 March 2004 09:45
		Re: BYTSCL and NAN keyword By: K. Bowman on Tue, 02 March 2004 09:26
		Re: BYTSCL and NAN keyword By: David Fanning on Tue, 02 March 2004 09:18
		Re: BYTSCL and NAN keyword By: Edd Edmondson on Tue, 02 March 2004 09:13
		Re: BYTSCL and NAN keyword By: K. Bowman on Wed, 03 March 2004 07:26
		Re: BYTSCL and NAN keyword By: Kenneth P. Bowman on Tue, 02 March 2004 17:44

Previous Topic:	Re: 2D-fit
Next Topic:	FastDL: An MPI interface for IDL

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Nov 29 15:07:37 PST 2025

Total time taken to generate the page: 0.00271 seconds