Re: st: manipulate ordering of rows matrices

[r.a.nobel@umail.leidenuniv.nl]

Yes I'm sorry. What is correct for me in this particular case, is that for
example for columns 3,4&5 (which are in fact dummy codings for a variable
consisting of 4 categories) the ones are sorted in an upward diagonal:

0 0 0
0 0 1
0 1 0
1 0 0

But I need this be done generally done as:

0 0 0
1 0 0
0 1 0
0 0 1

I hope this makes it more clear,

Alex



>Quoting Nick Cox <njcoxstata@gmail.com>:

> It would help if you explained the essence of the difference. What
> makes X2 sorted correctly?
>
> Nick
>
> On Tue, May 10, 2011 at 8:32 AM,  <r.a.nobel@umail.leidenuniv.nl> wrote:
>
> > I'm having trouble with the following in Mata:
> >
> > From a large matrix X, I need a new matrix with unique values/rows. As an
> > example, when I use :
> >
> > x1 = uniqrows(X)
> >
> > Then, the matrix given is:
> >
> > x1 = (
> > 1,0,0,0,0,0,0,0\
> > 1,0,0,0,1,0,0,0\
> > 1,0,0,1,0,0,0,0\
> > 1,0,1,0,0,0,0,0\
> > 1,1,0,0,0,0,0,0\
> > 1,1,0,0,1,0,0,1\
> > 1,1,0,1,0,0,1,0\
> > 1,1,1,0,0,1,0,0   )
> >
> > But what I need is:
> >
> > x2 = (
> > 1,0,0,0,0,0,0,0\
> > 1,0,1,0,0,0,0,0\
> > 1,0,0,1,0,0,0,0\
> > 1,0,0,0,1,0,0,0\
> > 1,1,0,0,0,0,0,0\
> > 1,1,1,0,0,1,0,0\
> > 1,1,0,1,0,0,1,0\
> > 1,1,0,0,1,0,0,1   )
> >
> > Of course, this can be done easily for just this matrix, but the problem is
> that
> > I need the correct matrix as part of an estimation command (thus, the
> matrices
> > can be different everytime, dependent on the dataset at hand), so I need a
> sort
> > of general way to ensure the matrix  is always sorted correctly. However,
> I've
> > not find a working strategy yet.
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>
>




*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/