| finding exclusive elements between two not-quite identical arrays [message #89740] |
Tue, 25 November 2014 22:51  |
havok2063
Messages: 24
Registered: December 2012
|
Junior Member |
|
|
So I know how to find elements in one array that are not in a second, when both arrays have identical elements.
A = [0,2,3,4]
B = [0,1,2,3,4]
print, where(histogram(A, omin=om) eq 0 and histogram(B,min=om) ne 0)+om
1
Now I want to do the same thing, but with two arrays containing integers that aren't quite identical in each one. Some of the elements can be off by +- 1. So
A = [11, 19, 40]
B = [10, 20, 30, 40]
Doing the above should return element index 2 (30) in B that is not in A, but I don't know how to do this. Any ideas?
Thanks, Brian
|
|
|
|
| Re: finding exclusive elements between two not-quite identical arrays [message #89741 is a reply to message #89740] |
Wed, 26 November 2014 03:05  |
Nikola
Messages: 53
Registered: November 2009
|
Member |
|
|
If the arrays are small and there is no other restriction on using loops, you can do it like this:
a = [11, 19, 40]
b = [10, 20, 30, 40]
c = b
margin = 5
FOR i = 0, N_ELEMENTS(b)-1 do c[i] = TOTAL(ABS((b[i]-a)) LT margin) EQ 0
PRINT, WHERE(c NE 0)
2
a = [11, 23, 40]
b = [10, 20, 30, 40]
c = b
margin = 5
FOR i = 0, N_ELEMENTS(b)-1 do c[i] = TOTAL(ABS((b[i]-a)) LT margin) EQ 0
PRINT, WHERE(c NE 0)
2
a = [11, 23, 40]
b = [10, 20, 30, 40]
c = b
margin = 2
FOR i = 0, N_ELEMENTS(b)-1 do c[i] = TOTAL(ABS((b[i]-a)) LT margin) EQ 0
PRINT, WHERE(c NE 0)
1 2
Where margin obviously specify how close elements of A and B should be to consider them matching. I don't see a quick solution to do this with histograms especially if b does not have to be equispaced. The method above work well in that case as well:
a = [11, 35, 40]
b = [10, 17, 33, 40]
c = b
margin = 5
FOR i = 0, N_ELEMENTS(b)-1 do c[i] = TOTAL(ABS((b[i]-a)) LT margin) EQ 0
PRINT, WHERE(c NE 0)
1
On Wednesday, November 26, 2014 6:51:19 AM UTC, Brian Cherinka wrote:
> So I know how to find elements in one array that are not in a second, when both arrays have identical elements.
>
> A = [0,2,3,4]
> B = [0,1,2,3,4]
>
> print, where(histogram(A, omin=om) eq 0 and histogram(B,min=om) ne 0)+om
> 1
>
> Now I want to do the same thing, but with two arrays containing integers that aren't quite identical in each one. Some of the elements can be off by +- 1. So
>
> A = [11, 19, 40]
> B = [10, 20, 30, 40]
>
> Doing the above should return element index 2 (30) in B that is not in A, but I don't know how to do this. Any ideas?
>
> Thanks, Brian
|
|
|
|
| Re: finding exclusive elements between two not-quite identical arrays [message #89742 is a reply to message #89740] |
Wed, 26 November 2014 03:37 |
Heinz Stege
Messages: 189
Registered: January 2003
|
Senior Member |
|
|
Hi Brian,
here is a quick draft:
A = [11, 19, 40]
B = [10, 20, 30, 40]
ii=value_locate(a,b) ; needs a to be sorted
delta=1
jj=where(b-a[ii] gt delta and a[ii+1]-b gt delta,count) ; needs the
; compile option strictarrsubs NOT to be set
if count ge 1 then print,'Part II: ',b[jj]
kk=where(a[0]-b gt delta,count) ; use min(a) if a is not sorted
if count ge 1 then print,'Part I: ',b[kk]
Good luck, Heinz
|
|
|
|
| Re: finding exclusive elements between two not-quite identical arrays [message #89743 is a reply to message #89740] |
Wed, 26 November 2014 06:15 |
Phillip Bitzer
Messages: 223
Registered: June 2006
|
Senior Member |
|
|
On Wednesday, November 26, 2014 12:51:19 AM UTC-6, Brian Cherinka wrote:
> So I know how to find elements in one array that are not in a second, when both arrays have identical elements.
>
> A = [0,2,3,4]
> B = [0,1,2,3,4]
>
You can also try Craig's CMSET_OP:
http://www.physics.wisc.edu/~craigm/idl/arrays.html
|
|
|
|
| Re: finding exclusive elements between two not-quite identical arrays [message #89744 is a reply to message #89742] |
Wed, 26 November 2014 11:59 |
havok2063
Messages: 24
Registered: December 2012
|
Junior Member |
|
|
On Wednesday, November 26, 2014 6:37:34 AM UTC-5, Heinz Stege wrote:
> Hi Brian,
>
> here is a quick draft:
>
> A = [11, 19, 40]
> B = [10, 20, 30, 40]
> ii=value_locate(a,b) ; needs a to be sorted
> delta=1
> jj=where(b-a[ii] gt delta and a[ii+1]-b gt delta,count) ; needs the
> ; compile option strictarrsubs NOT to be set
> if count ge 1 then print,'Part II: ',b[jj]
> kk=where(a[0]-b gt delta,count) ; use min(a) if a is not sorted
> if count ge 1 then print,'Part I: ',b[kk]
>
> Good luck, Heinz
Thanks guys. The solution offered by Heinz works best for me. I was also looking for a simple non-loop solution. I like Craig's functions, but I think CMSET_OP only works on identical elements. Thanks again everyone.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
