| Re: Finding Common Elements in Two Arrays [message #2104] |
Thu, 02 June 1994 07:23 |
zawodny
Messages: 121
Registered: August 1992
|
Senior Member |
|
|
In article <2skpl4$qr5@reznor.larc.nasa.gov> zawodny@arbd0.larc.nasa.gov (Joseph M Zawodny) writes:
> In article <1JUN199416454238@stars.gsfc.nasa.gov>, kucera@stars.gsfc.nasa.gov (Terry Kucera) writes:
>> I'm looking for a quick way to compare two arrays in IDL, A and B,
>> and determine which elements of B are also in A,
>> so if:
>> A=[2,1,3,5,3,8,2,5]
>> B=[3,4,2,8,7,8]
>> I would get [0,2,3,5], because 3, 2, and 8 are in A as well as B.
>>
>> I can do this with loops, but that takes too long for big arrays. Does anyone
>> have a way to do this using array functions or perhaps an external routine?
>> Terry Kucera
>> kucera@stars.gsfc.nasa.gov
>
> If you are dealing strictly with integers, I'd use histogram on both
> vectors, multiply them together and then use where to find non-zero values.
>
> Hist_a = histogram(a,max=maximum_expected_value)
> Hist_b = histogram(b,max=maximum_expected_value)
> Inter = where(Hist_a*Hist_b)
>
> and Inter should then contain the array you seek.
No, Sorry, I mis read your post. Inter tells you the value of the elemnts that
are both arrays, not which elements are the same. Hist_b does, however,
contain this information. You will also have to use the REVERSE_INDICES keyword
in the second histogram call. Basically you'd have to use Inter and Hist_b
jointly to construct an array of subscripts.
--
Joseph M. Zawodny (KO4LW) NASA Langley Research Center
Internet: zawodny@arbd0.larc.nasa.gov MS-475, Hampton VA, 23681-0001
Packet: ko4lw@n4hog.va.usa
|
|
|
|
| Re: Finding Common Elements in Two Arrays [message #2106 is a reply to message #2104] |
Thu, 02 June 1994 07:17  |
zawodny
Messages: 121
Registered: August 1992
|
Senior Member |
|
|
In article <1JUN199416454238@stars.gsfc.nasa.gov>, kucera@stars.gsfc.nasa.gov (Terry Kucera) writes:
> I'm looking for a quick way to compare two arrays in IDL, A and B,
> and determine which elements of B are also in A,
> so if:
> A=[2,1,3,5,3,8,2,5]
> B=[3,4,2,8,7,8]
> I would get [0,2,3,5], because 3, 2, and 8 are in A as well as B.
>
> I can do this with loops, but that takes too long for big arrays. Does anyone
> have a way to do this using array functions or perhaps an external routine?
> Terry Kucera
> kucera@stars.gsfc.nasa.gov
If you are dealing strictly with integers, I'd use histogram on both
vectors, multiply them together and then use where to find non-zero values.
Hist_a = histogram(a,max=maximum_expected_value)
Hist_b = histogram(b,max=maximum_expected_value)
Inter = where(Hist_a*Hist_b)
and Inter should then contain the array you seek.
--
Joseph M. Zawodny (KO4LW) NASA Langley Research Center
Internet: zawodny@arbd0.larc.nasa.gov MS-475, Hampton VA, 23681-0001
Packet: ko4lw@n4hog.va.usa
|
|
|
|
| Re: Finding Common Elements in Two Arrays [message #2109 is a reply to message #2106] |
Thu, 02 June 1994 05:04 |
dutch
Messages: 8
Registered: June 1992
|
Junior Member |
|
|
In article <1JUN199416454238@stars.gsfc.nasa.gov>, kucera@stars.gsfc.nasa.gov (Terry Kucera) writes:
>> I'm looking for a quick way to compare two arrays in IDL, A and B,
>> and determine which elements of B are also in A,
>> so if:
>> A=[2,1,3,5,3,8,2,5]
>> B=[3,4,2,8,7,8]
>> I would get [0,2,3,5], because 3, 2, and 8 are in A as well as B.
>>
>> I can do this with loops, but that takes too long for big arrays. Does anyone
>> have a way to do this using array functions or perhaps an external routine?
>> Terry Kucera
>> kucera@stars.gsfc.nasa.gov
>>
Perhaps the following function is what you are looking for. As it is currently
written, the function returns the VALUES of repeated elements, rather than
the INDICES into the first vector (as in your example), but this is a trivial
modification, if necessary!.
The only hiccup, is what to return if there are no repeated elements,
in my case I have chosen to return the value -999. This function is a rewrite
of a similar function in MATLAB. (in MATLAB, the function can return the empty
set as a null vector (with zero elements). In IDL, the function could be modified to return a second argument, which indicates the number of elements in the intersection.
Rough benchmarks on our Vax 4090 are as follows:
2 random vectors each of length 10,000 elements - about 1 second
2 random vectors each of length 100,000 elements - about 12 seconds
I expect this is considerably better than alternative solutions using one or
more FOR loops!.
---------- 8< start: intersection.pro >8 -------------
FUNCTION intersection,a,b
;+
; general purpose routine to return the intersection
; (i.e. common elements) of two vectors. Uses USERLIB
; routine UNIQ to obtain unique elements of a,b
; Written by M.J.Dutch June-1994.
; Centre de Recherches en Physique des Plasmas, EPFL, Switzerland
;-
ab=[a(uniq(a,sort(a))),b(uniq(b,sort(b)))] ; combine unique elements of a,b
ab=ab(sort(ab)) ; sort the combined vector
nab=n_elements(ab)
diff=ab(1:nab-1)-ab(0:nab-2)
ind=where(diff eq 0) ; find repeated elements
if (ind(0) ne -1) then return,ab(ind) else return,-999
end
---------- 8< end: intersection.pro >8 -------------
Enjoy!
-- Michael
###########################################################
# Dr. Michael Dutch email: dutch@eltcv1.epfl.ch #
# Centre de Recherches en Physique des Plasmas #
# Ecole Polytechnique Federale de Lausanne #
# 21 Ave des Bains Aussie.Abroad #
# CH-1007 Lausanne, SWITZERLAND ,--_|\ #
#---------------------------------------- / \ #
# I'd rather have a full bottle in front \_.-*._/ #
# of me than a full frontal lobotomy. v #
###########################################################
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
