Indexing where all entries are false

[Richard Bovey]

The result of indexing a vector with logicals has changed from 3.0.0 to 3.0.1. I believe that the 3.0.1 behaviour maybe considered incorrect as it is inconsistent with Matlab, while 3.0.0 was consistent with Matlab.

In 3.0.1 if all the entries in the index are false a matrix with size 0x0 is returned, while 3.0.1 returns 1x0 or 0x1 depending on the shape of the input vector.

I'm using Windows XP.

Below are some example sessions in 3.0.1 and 3.0.0.

------------------

>> OCTAVE_VERSION
ans = 3.0.1
>> x = [1 1 1]
x =

   1   1   1

>> x([false false false])
ans = [](0x0)

------------------

>> OCTAVE_VERSION
ans = 3.0.0
>> x = [1 1 1]
x =

   1   1   1

>> x([false false false])
ans = [](1x0)

Regards,
Richard

[David Bateman]

Richard Bovey wrote:

> The result of indexing a vector with logicals has changed from 3.0.0 to
> 3.0.1. I believe that the 3.0.1 behaviour maybe considered incorrect as it
> is inconsistent with Matlab, while 3.0.0 was consistent with Matlab.
>
> In 3.0.1 if all the entries in the index are false a matrix with size 0x0 is
> returned, while 3.0.1 returns 1x0 or 0x1 depending on the shape of the input
> vector.
>
> I'm using Windows XP.
>
>  

The cause seems to be the change

http://velveeta.che.wisc.edu/cgi-bin/hgwebdir.cgi/octave/diff/85da2ab0c6fd/liboctave/idx-vector.cc

where if the length of an index vector is zero when created from an
boolNDArray then the dimensions of the idnex vector are incorrectly set
to dim_vector(0,0). The changelog entry for this change is

2008-03-07  John W. Eaton  <jwe@...>

        * idx-vector.cc (IDX_VEC_REP::idx_vector_rep (bool),
        IDX_VEC_REP::idx_vector_rep (const boolNDArray&)):
        Simply perform the equivalent of "find" on the bool argument here,
        set one_zero to 0 and orig_dims to size of resulting index vector.
        (IDX_VEC_REP::freeze): Don't call maybe_convert_one_zero_to_idx
here.

The changest log is marked as "logical indexing compatibility fixes
<http://velveeta.che.wisc.edu/cgi-bin/hgwebdir.cgi/octave/rev/5cd053b8d9cd>"
and so we'd have to be careful not to just back out this change but also
address the original issue. The original problem is discuss here

http://www.nabble.com/Incompatibility-with-matlab-logical-indexing-to15908186.html

We therefore also need to check the following

>> a = [1 2; 3 4]

a =

     1     2
     3     4

>> a([1, 1] == 1)

ans =

     1     3

>> a([1, 1] == 0)

ans =

   Empty matrix: 1-by-0

>> a([1; 1] == 0)

ans =

   Empty matrix: 0-by-1

which is the output from MatlabR2007b. I suppose the fix is then the
attached which seems to address both your problem, the above and the
original problem.

Regards
David

--
David Bateman                                David.Bateman@...
Motorola Labs - Paris                        [phone] (Ph)
Parc Les Algorithmes, Commune de St Aubin    [phone] (Mob)
91193 Gif-Sur-Yvette FRANCE                  [phone] (Fax)

The information contained in this communication has been classified as:

[x] General Business Information
[ ] Motorola Internal Use Only
[ ] Motorola Confidential Proprietary


# HG changeset patch
# User David Bateman <dbateman@...>
# Date 1210693535 -7200
# Node ID e4affe5a2f978db78f6efbd705a79304b3bdc3d9
# Parent  9210dcd13292ed5f686ea2a926aa77bda7d0fcde
all false logical indexing fix

diff --git a/liboctave/ChangeLog b/liboctave/ChangeLog
--- a/liboctave/ChangeLog
+++ b/liboctave/ChangeLog
@@ -1,3 +1,9 @@ 2008-05-06  David Bateman  <dbateman@fre
[phone]  John W. Eaton  <jwe@...>
+
+ * idx-vector.cc (IDX_VEC_REP::idx_vector_rep (const boolNDArray&)):
+ If len is zero size the index vector in the same manner as if len
+ is not zero.
+
 2008-05-06  David Bateman  <dbateman@...>
 
  * Array.cc (Array<T> Array<T>::transpose () const): Modify for tiled
diff --git a/liboctave/idx-vector.cc b/liboctave/idx-vector.cc
--- a/liboctave/idx-vector.cc
+++ b/liboctave/idx-vector.cc
@@ -270,11 +270,13 @@ IDX_VEC_REP::idx_vector_rep (const boolN
     range(0), initialized (0), frozen (0),
     colon_equiv_checked (0), colon_equiv (0), orig_dims ()
 {
+  dim_vector dv = bnda.dims ();
+
+  orig_dims = ((dv.length () == 2 && dv(0) == 1)
+       ? dim_vector (1, len) : orig_dims = dim_vector (len, 1));
+
   if (len == 0)
-    {
-      orig_dims = dim_vector (0, 0);
-      initialized = 1;
-    }
+    initialized = 1;
   else
     {
       data = new octave_idx_type [len];
@@ -284,11 +286,6 @@ IDX_VEC_REP::idx_vector_rep (const boolN
       for (octave_idx_type i = 0, k = 0; i < ntot && k < len; i++)
  if (bnda.elem (i))
   data[k++] = i;
-
-      dim_vector dv = bnda.dims ();
-
-      orig_dims = ((dv.length () == 2 && dv(0) == 1)
-   ? dim_vector (1, len) : orig_dims = dim_vector (len, 1));
 
       init_state ();
     }

_______________________________________________
Bug-octave mailing list
Bug-octave@...
https://www.cae.wisc.edu/mailman/listinfo/bug-octave

[John W. Eaton]

On 13-May-2008, David Bateman wrote:

| Richard Bovey wrote:
| > The result of indexing a vector with logicals has changed from 3.0.0 to
| > 3.0.1. I believe that the 3.0.1 behaviour maybe considered incorrect as it
| > is inconsistent with Matlab, while 3.0.0 was consistent with Matlab.
| >
| > In 3.0.1 if all the entries in the index are false a matrix with size 0x0 is
| > returned, while 3.0.1 returns 1x0 or 0x1 depending on the shape of the input
| > vector.
| >
| > I'm using Windows XP.
| >
| >  
| The cause seems to be the change
|
| http://velveeta.che.wisc.edu/cgi-bin/hgwebdir.cgi/octave/diff/85da2ab0c6fd/liboctave/idx-vector.cc
|
| where if the length of an index vector is zero when created from an
| boolNDArray then the dimensions of the idnex vector are incorrectly set
| to dim_vector(0,0). The changelog entry for this change is
|
| 2008-03-07  John W. Eaton  <jwe@...>
|
|         * idx-vector.cc (IDX_VEC_REP::idx_vector_rep (bool),
|         IDX_VEC_REP::idx_vector_rep (const boolNDArray&)):
|         Simply perform the equivalent of "find" on the bool argument here,
|         set one_zero to 0 and orig_dims to size of resulting index vector.
|         (IDX_VEC_REP::freeze): Don't call maybe_convert_one_zero_to_idx
| here.
|
| The changest log is marked as "logical indexing compatibility fixes
| <http://velveeta.che.wisc.edu/cgi-bin/hgwebdir.cgi/octave/rev/5cd053b8d9cd>"
| and so we'd have to be careful not to just back out this change but also
| address the original issue. The original problem is discuss here
|
| http://www.nabble.com/Incompatibility-with-matlab-logical-indexing-to15908186.html
|
| We therefore also need to check the following
|
| >> a = [1 2; 3 4]
|
| a =
|
|      1     2
|      3     4
|
| >> a([1, 1] == 1)
|
| ans =
|
|      1     3
|
| >> a([1, 1] == 0)
|
| ans =
|
|    Empty matrix: 1-by-0
|
| >> a([1; 1] == 0)
|
| ans =
|
|    Empty matrix: 0-by-1
|
| which is the output from MatlabR2007b. I suppose the fix is then the
| attached which seems to address both your problem, the above and the
| original problem.

I applied this patch, but replaced my name with yours in the ChangeLog
entry.

Thanks,

jwe
_______________________________________________
Bug-octave mailing list
Bug-octave@...
https://www.cae.wisc.edu/mailman/listinfo/bug-octave

[David Bateman]

John W. Eaton wrote:
> I applied this patch, but replaced my name with yours in the ChangeLog
> entry.
>
>  
I think this fix should also be applied to the 3.0.x tree as well, but
is only applied to the 3.1.x tree..

Cheers
David


--
David Bateman                                David.Bateman@...
Motorola Labs - Paris                        [phone] (Ph)
Parc Les Algorithmes, Commune de St Aubin    [phone] (Mob)
91193 Gif-Sur-Yvette FRANCE                  [phone] (Fax)

The information contained in this communication has been classified as:

[x] General Business Information
[ ] Motorola Internal Use Only
[ ] Motorola Confidential Proprietary

_______________________________________________
Bug-octave mailing list
Bug-octave@...
https://www.cae.wisc.edu/mailman/listinfo/bug-octave

[John W. Eaton]

On 15-May-2008, David Bateman wrote:

| John W. Eaton wrote:
| > I applied this patch, but replaced my name with yours in the ChangeLog
| > entry.
| >
| >  
| I think this fix should also be applied to the 3.0.x tree as well, but
| is only applied to the 3.1.x tree..

OK, I applied it to the release-3-0-x branch.

Thanks,

jwe
_______________________________________________
Bug-octave mailing list
Bug-octave@...
https://www.cae.wisc.edu/mailman/listinfo/bug-octave