Re: uniq -c

[Mike Miller]

On Fri, 27 Aug 2004, John W. Eaton wrote:

> On 27-Aug-2004, Mike Miller <address@hidden> wrote:
>
> | Is there a way to count repetitions of a value in a vector?  In unix/linux
> | we could use "uniq -c" to count consecutive identical lines.  Is there
> | something analogous in Octave?  If not, can one of you come up with a
> | clever scheme for doing this within Octave?  (I've been on this list long
> | enough to know that the answer is yes to that last question!)
> |
> | I really just need the counts and not a list of the values being counted.
> | Hmm... I just realized that if I had an ordinary 'uniq' function, I could
> | get the "uniq -c" this way (for column vector x):
> |
> | [NN,XX] = hist(x,uniq(x))
> |
> | Then [NN,XX] would be the "uniq -c" output and XX would actually equal
> | uniq(x).
>
> Octave has
>
>  create_set
>  intersection
>  union


Thanks!  So create_set(X) essentially does "vec(X) | sort | uniq" (pardon 
my use of unixy pipes) on the matrix X.  I needed the sorting, and this 
happened to give me what I needed:

[NN,XX]=hist(x,create_set(x)); NN


We might consider having a uniq(X) operation that doesn't sort the data. 
We can always do uniq(sort(X)), as in UNIX/Linux.  If uniq operated on 
rows of a matrix, that would work for me, but it would probably be most 
useful with column vectors.

Thanks again.

Mike



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------