comp.lang.idl-pvwave archive: archive » Re: Vector comparison.

Home » Public Forums » archive » Re: Vector comparison.

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

Re: Vector comparison. [message #37059 is a reply to message #37058]

Thu, 20 November 2003 07:03

tam
Messages: 48
Registered: February 2000

Member

David Fanning wrote:

> Reimar Bauer writes:
>
>
>> P.S. If we have all read it as carefully as mentioned what would we have
>> answered. Pavel will uses loops as well, we boths have not read
>> carefully enough, David hasn't replied.
>
>
> No, David was interested in the *answer*. He didn't
> have a solution of his own. :-)
>
> Cheers,
>
> David

How about the following, which handles duplicates in both arrays
and uses no loops.

; Find all of the indices of the elements
; in the first array that are matched in the second
; array
function match, a, b

m = n_elements(a)
n = n_elements(b)

; Create an array with each possible pairing of the first
; and last elements

cmp = replicate(a[0], 2, m*n)

ind = lindgen(m*n)
ind0 = ind mod m
ind1 = ind / m

; Just lay out the 'a' array multiple times.
cmp[0,*] = a[ind0]

; Repeat each element of the 'b' array m times so that
; we get each a element paired with each b element.
cmp[1,*] = b[ind1]

; Now find all of the matches.
w = where(cmp[0,*] eq cmp[1,*])

if (w[0] eq -1) then begin
return, w
endif else begin

; Handle multiples in the 'b' set.
; If the elements in 'b' are guranteed
; to be unique then we can just return 'w mod m'

h = histogram(w mod m, min=0)
return, where(h ne 0)
endelse
end

Have I missed something here... I think this would
be reasonably efficient. Probably don't want to
have separate ind, ind0, and ind1 arrays, but I thought
that might show the algorithm more clearly.

cmp[*,*] = [[a[lindgen(m*n) mod m][b[lindgen(m*n)/m]]

is not the kind of thing I can follow!

Regards,
Tom McGlynn

Report message to a moderator

[Message index]

		Re: Vector comparison. By: JD Smith on Thu, 20 November 2003 11:38
		Re: Vector comparison. By: tam on Thu, 20 November 2003 10:55
		Re: Vector comparison. By: David Fanning on Thu, 20 November 2003 09:55
		Re: Vector comparison. By: Craig Markwardt on Thu, 20 November 2003 09:35
		Re: Vector comparison. By: tam on Thu, 20 November 2003 09:24
		Re: Vector comparison. By: David Fanning on Thu, 20 November 2003 08:52
		Re: Vector comparison. By: Chris[1] on Thu, 20 November 2003 08:08
		Re: Vector comparison. By: David Fanning on Thu, 20 November 2003 08:08
		Re: Vector comparison. By: R.Bauer on Thu, 20 November 2003 07:11
		Re: Vector comparison. By: tam on Thu, 20 November 2003 07:03
		Re: Vector comparison. By: hunter on Thu, 20 November 2003 06:49
		Re: Vector comparison. By: David Fanning on Thu, 20 November 2003 05:40
		Re: Vector comparison. By: R.Bauer on Thu, 20 November 2003 04:02
		Re: Vector comparison. By: Craig Markwardt on Thu, 20 November 2003 02:19
		Re: Vector comparison. By: Chris Lee on Thu, 20 November 2003 02:09
		Re: Vector comparison. By: R.Bauer on Thu, 20 November 2003 01:01
		Re: Vector comparison. By: Craig Markwardt on Thu, 20 November 2003 00:23
		Re: Vector comparison. By: David Fanning on Wed, 19 November 2003 18:05
		Re: Vector comparison. By: R.Bauer on Wed, 19 November 2003 17:50
		Re: Vector comparison. By: Craig Markwardt on Wed, 19 November 2003 15:48
		Re: Vector comparison. By: Pavel Romashkin on Wed, 19 November 2003 16:09

Previous Topic:	Re: IDLWAVE takes over
Next Topic:	Byte Swapping

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Oct 11 05:13:59 PDT 2025

Total time taken to generate the page: 1.12188 seconds