|
|
| Re: rotate a large data volume [message #45426 is a reply to message #45421] |
Thu, 08 September 2005 05:53  |
peter.albert@gmx.de
Messages: 108
Registered: July 2005
|
Senior Member |
|
|
> Ooops!! I guess that's rotating about the x-axis isn't it?
Fine, so that's settled :-)
Well, in this case I am a bit stuck, as you can't read in all the data
at once in oder to do some fancy things with histogram. So I guess you
are right, clever reading from disk is necessary, and I doubt it's
something you can easily solve with IDL. Well, I'll think over it over
the week-end (starting today :-)) and will hope that someone else comes
with a good solution ...
Best regards,
Peter
|
|
|
|
| Re: rotate a large data volume [message #45434 is a reply to message #45426] |
Thu, 08 September 2005 00:35 |
Jeff N.
Messages: 120
Registered: April 2005
|
Senior Member |
|
|
Ooops!! I guess that's rotating about the x-axis isn't it?
Whew....sorry, you can tell from all my posting that its late and I've
been working WAY to long today....
|
|
|
|
| Re: rotate a large data volume [message #45435 is a reply to message #45434] |
Thu, 08 September 2005 00:11 |
Jeff N.
Messages: 120
Registered: April 2005
|
Senior Member |
|
|
I suppose I should also point out that I can already make this
work....in fact the person that asked me for help already had it
working....I just need to make it work in a way that doesn't take 4
days now :)
I can post you two versions of code if it'll help: one is the original
code by the person that asked me for help, and the second is what I
came up with using ENVI tiling routines.
I think that the problem is that i'm trying to move around columns of
data, which aren't in contiguous locations on disk, and its the read
times that are killing me.
|
|
|
|
| Re: rotate a large data volume [message #45436 is a reply to message #45435] |
Thu, 08 September 2005 00:05 |
Jeff N.
Messages: 120
Registered: April 2005
|
Senior Member |
|
|
Sure Peter,
Here's an example with a 3x3x3 cube that I'd already made up so that I
could get a handle on problem myself (I'm helping out a friend here,
its not really my problem....though I guess it has become that way I
suppose :) )
These three input bands (BSQ interleave):
in0
9 4 5
2 8 6
2 6 4
in1
2 8 1
3 5 2
1 1 6
in2
1 7 4
7 6 2
3 8 5
should give a 1st output band of
5 1 4
6 2 2
4 6 5
Hope this helps!
Jeff
|
|
|
|
| Re: rotate a large data volume [message #45437 is a reply to message #45436] |
Wed, 07 September 2005 23:01 |
peter.albert@gmx.de
Messages: 108
Registered: July 2005
|
Senior Member |
|
|
Hi Jeff,
I am most probably getting something wrong here, but if I assume that
each tiff file represents one x/y layer, then rotating about the z axis
would just mean that you have to rotate each individual file, which you
could (on a Unix machine) easily do with e.g. convert without bothering
IDL.
However, I guess there is more to it, so I'd appreciate if you could
clarify the problem with e.g. a 2x2x3 example.
Regards,
Peter
|
|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
