[Octave-bug-tracker] [bug #59191] Using (end+1) on an empty matrix refers causes unexpected padding with zeros

[anonymous]

URL:
  <https://savannah.gnu.org/bugs/?59191>

                 Summary: Using (end+1) on an empty matrix refers causes
unexpected padding with zeros
                 Project: GNU Octave
            Submitted by: None
            Submitted on: Sun 27 Sep 2020 02:00:12 PM UTC
                Category: Interpreter
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Inaccurate Result
                  Status: None
             Assigned to: None
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
                 Release: dev
         Discussion Lock: Any
        Operating System: GNU/Linux

    _______________________________________________________

Details:

Octave version 7.0.0 with hg id 901a92f45ff0 on 64-bit Linux

When augmenting an array by using (end+1) as the subscript on an lvalue, it
works properly for one or two dimensions, but for three or more dimensions it
causes the array to be unexpectedly pre-padded with an extra layer of zeros.
It appears that (end+1) for an empty matrix refers to subscript value 1 for
one and two dimensions, but subscript value 2 for three or more dimensions. 

Once this behavior is known, it can certainly be worked around by the user. It
is being reported in case the behavior was not intended.


octave:7> lst = []; lst(end+1,:) = pi(1,4), size(lst)
lst =
   3.1416   3.1416   3.1416   3.1416

ans =
   1   4

octave:8> lst = []; lst(:,end+1) = pi(1,4), size(lst)
lst =
   3.1416
   3.1416
   3.1416
   3.1416

ans =
   4   1

octave:9> lst = []; lst(:,:,end+1) = pi(4,4), size(lst)
lst =
ans(:,:,1) =
   0   0   0   0
   0   0   0   0
   0   0   0   0
   0   0   0   0
ans(:,:,2) =
   3.1416   3.1416   3.1416   3.1416
   3.1416   3.1416   3.1416   3.1416
   3.1416   3.1416   3.1416   3.1416
   3.1416   3.1416   3.1416   3.1416

ans =
   4   4   2
octave:10> lst = []; lst(:,:,:,end+1) = pi(4,4), size(lst)
lst =
ans(:,:,1,1) =
   0   0   0   0
   0   0   0   0
   0   0   0   0
   0   0   0   0
ans(:,:,1,2) =
   3.1416   3.1416   3.1416   3.1416
   3.1416   3.1416   3.1416   3.1416
   3.1416   3.1416   3.1416   3.1416
   3.1416   3.1416   3.1416   3.1416

ans =
   4   4   1   2






    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?59191>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/