Thread Subject:

How to use "Unique"

Hello,

I am having trouble figuring out how to use "unique" to find the index of identical elements in a single column cell array. As an example, I have a cell array with the following data:

A =
    ''231234242800100043062.68'
    ''231234242800200043208.86'
    ''231234242800400042964.91'
    ''231234242800500038861.19'
    ''231234242800600010255.99'
    ''231234242800900054183.5'
    ''231234242801100077938.9'
    ''231234242801200049682.26'
    ''23123424800900054183.5'
    ''23123424800900054183.5'
    ''23123424GHA0200020256.36'

I have been trying to use unique to find the index of identical elements. In this case, the third and second to last values are equal to one another. I need to figure out a way to output the indices 9 & 10. I have tried using [C1, ia1, ib1] = unique(A,'first'), but still do not understand how "ib1" can be used to find the indices 9 & 10.

Also, the order is crucial. The order shown in the example cell array cannot change. I realize this may be a question for the help file, but after looking at it I still do not fully understand so I am hoping someone could help. Thanks.

Kevin

On 11/12/2012 11:10 AM, Kevin Ellis wrote:
...
> I am having trouble figuring out how to use "unique" to find the index
> of identical elements in a single column cell array. ...
...
> Also, the order is crucial. The order shown in the example cell array
> cannot change. I realize this may be a question for the help file, but
> after looking at it I still do not fully understand so I am hoping
> someone could help. Thanks.
...

unique() will give you the values that are, well, unique to the set.
Once you have them you can then use findstr() and/or ismember() to
locate the locations of those values in the original vector.

Or, histc for numeric values for the bins obtained from unique() but
your data are/must be character in the above example...

--

In article <k7rai0$n1i$1@newscl01ah.mathworks.com>,
 "Kevin Ellis" <kevin.ellis86@gmail.com> wrote:

> Hello,
>
> I am having trouble figuring out how to use "unique" to find the index of
> identical elements in a single column cell array. As an example, I have a
> cell array with the following data:
>
> A =
> ''231234242800100043062.68'
> ''231234242800200043208.86'
> ''231234242800400042964.91'
> ''231234242800500038861.19'
> ''231234242800600010255.99'
> ''231234242800900054183.5'
> ''231234242801100077938.9'
> ''231234242801200049682.26'
> ''23123424800900054183.5'
> ''23123424800900054183.5'
> ''23123424GHA0200020256.36'
>
> I have been trying to use unique to find the index of identical elements. In
> this case, the third and second to last values are equal to one another. I
> need to figure out a way to output the indices 9 & 10. I have tried using
> [C1, ia1, ib1] = unique(A,'first'), but still do not understand how "ib1" can
> be used to find the indices 9 & 10.
>
> Also, the order is crucial. The order shown in the example cell array cannot
> change. I realize this may be a question for the help file, but after looking
> at it I still do not fully understand so I am hoping someone could help.
> Thanks.
>
> Kevin

Try this:

[~,~,ib1] = unique(A);
j = find(histc(ib1,1:max(ib1)) > 1);
dupes = find(ismember(ib1,j));

It will find ALL indices of items in A that are duplicated, even if
there is more than one such unique item, i.e., if

A = {'a','a','b','c','c'}

then dupes will be [1 2 4 5];

--
Doug Schwarz
dmschwarz&ieee,org
Make obvious changes to get real email address.

Doug Schwarz <see@sig.for.address.edu> wrote in message <see-8B363E.12420112112012@news.eternal-september.org>...
> In article <k7rai0$n1i$1@newscl01ah.mathworks.com>,
> "Kevin Ellis" <kevin.ellis86@gmail.com> wrote:
>
> > Hello,
> >
> > I am having trouble figuring out how to use "unique" to find the index of
> > identical elements in a single column cell array. As an example, I have a
> > cell array with the following data:
> >
> > A =
> > ''231234242800100043062.68'
> > ''231234242800200043208.86'
> > ''231234242800400042964.91'
> > ''231234242800500038861.19'
> > ''231234242800600010255.99'
> > ''231234242800900054183.5'
> > ''231234242801100077938.9'
> > ''231234242801200049682.26'
> > ''23123424800900054183.5'
> > ''23123424800900054183.5'
> > ''23123424GHA0200020256.36'
> >
> > I have been trying to use unique to find the index of identical elements. In
> > this case, the third and second to last values are equal to one another. I
> > need to figure out a way to output the indices 9 & 10. I have tried using
> > [C1, ia1, ib1] = unique(A,'first'), but still do not understand how "ib1" can
> > be used to find the indices 9 & 10.
> >
> > Also, the order is crucial. The order shown in the example cell array cannot
> > change. I realize this may be a question for the help file, but after looking
> > at it I still do not fully understand so I am hoping someone could help.
> > Thanks.
> >
> > Kevin
>
> Try this:
>
> [~,~,ib1] = unique(A);
> j = find(histc(ib1,1:max(ib1)) > 1);
> dupes = find(ismember(ib1,j));
>
> It will find ALL indices of items in A that are duplicated, even if
> there is more than one such unique item, i.e., if
>
> A = {'a','a','b','c','c'}
>
> then dupes will be [1 2 4 5];
>
> --
> Doug Schwarz
> dmschwarz&ieee,org
> Make obvious changes to get real email address.

Thanks for the quick response. That worked perfectly.

Kevin