[Octave-bug-tracker] [bug #56950] NaN/Inf values in ind2rgb

[Rik]

Update of bug #56950 (project octave):

              Item Group:        Unexpected Error => Matlab Compatibility   

    _______________________________________________________

Follow-up Comment #1:

Octave is generally stricter about checking its inputs than Matlab.  In this
case, why would an indexed image have a NaN value?

Even if there is a use case for have a NaN value in an indexed image, it is a
rare occurrence.  For most users, a NaN is an error, and for the few users for
which it is not, I believe it is better for them to actively choose how they
want these values to be interpreted.  This is an easy one-line piece of code
as you demonstrated.  I am tempted to mark this as "Won't Fix" and have
programmers choose how they want such images to be interpreted.

    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?56950>

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