| Re: Empirical orthogonal function in IDL [message #82556] |
Fri, 21 December 2012 17:00 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
mkmvarma@gmail.com writes:
> Thanks, David. I will try to find the book. And one more question. Did you try to remove missing values from the data set?
I didn't have any missing data, since I was using
reanalysis data, but, yes, I would have removed it.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
Sepore ma de ni thue. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Empirical orthogonal function in IDL [message #82557 is a reply to message #82556] |
Fri, 21 December 2012 16:35  |
mkmvarma
Messages: 24
Registered: November 2007
|
Junior Member |
|
|
Thanks, David. I will try to find the book. And one more question. Did you try to remove missing values from the data set?
Thanks,
Mahesh
On Friday, December 21, 2012 2:43:48 PM UTC-8, David Fanning wrote:
> mkmvarma@gmail.com writes:
>
>
>
>> Thanks, David. I looked it up and understand your code except the last part of the code where PC of mode 1. The array 'pc[mode-1,*]' has elements 2592 but the array 'years' has only 21 elements. Could you please explain me how we can plot a timeseries plot using the principal component array.
>
>
>
> Well, that is the Wilks trick I mention in the article. I
>
> probably didn't understand it well enough four years ago
>
> when I wrote the article to explain it any better than
>
> I did. I certainly don't remember the trick now! I would,
>
> however, get over to the library and find that book. It
>
> is a good one!
>
>
>
> To prove to myself it worked, I programmed up the example
>
> the slow way (25 minutes or so) and the fast way (half
>
> a second). I do remember that the results were identical,
>
> so I guess I trusted the result, even if I didn't understand
>
> it completely. :-)
>
>
>
> Cheers,
>
>
>
> David
>
>
>
>
>
>
>
> --
>
> David Fanning, Ph.D.
>
> Fanning Software Consulting, Inc.
>
> Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
>
> Sepore ma de ni thue. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Empirical orthogonal function in IDL [message #82560 is a reply to message #82557] |
Fri, 21 December 2012 14:43 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
mkmvarma@gmail.com writes:
> Thanks, David. I looked it up and understand your code except the last part of the code where PC of mode 1. The array 'pc[mode-1,*]' has elements 2592 but the array 'years' has only 21 elements. Could you please explain me how we can plot a timeseries plot using the principal component array.
Well, that is the Wilks trick I mention in the article. I
probably didn't understand it well enough four years ago
when I wrote the article to explain it any better than
I did. I certainly don't remember the trick now! I would,
however, get over to the library and find that book. It
is a good one!
To prove to myself it worked, I programmed up the example
the slow way (25 minutes or so) and the fast way (half
a second). I do remember that the results were identical,
so I guess I trusted the result, even if I didn't understand
it completely. :-)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
Sepore ma de ni thue. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Empirical orthogonal function in IDL [message #82561 is a reply to message #82560] |
Fri, 21 December 2012 14:21 |
mkmvarma
Messages: 24
Registered: November 2007
|
Junior Member |
|
|
Thanks, David. I looked it up and understand your code except the last part of the code where PC of mode 1. The array 'pc[mode-1,*]' has elements 2592 but the array 'years' has only 21 elements. Could you please explain me how we can plot a timeseries plot using the principal component array.
Thanks,
Mahesh
|
|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
