On 11 May 2011 15:14, Thomas Weber <[email protected]> wrote: > On Tue, May 10, 2011 at 06:24:41PM -0500, Jordi Gutiérrez Hermoso wrote: >> On 10 May 2011 14:51, Thomas Weber <[email protected]> wrote: >> > On Tue, May 10, 2011 at 01:54:24PM +0200, Didier Raboud wrote: >> >> Package: octave-miscellaneous >> >> Version: 1.0.9-1+b1 >> >> Severity: minor >> >> >> >> Hi, >> >> >> >> octave-optim, in it 1.0.16-1 version, uploaded on 23. April, depends on >> >> octave-miscellaneous >= 1.0.11, so please upload it to allow the >> >> installation >> >> of octave-optim/unstable. >> > >> > From miscellaneous 1.0.11: >> > [map] >> > ***** test >> > assert(map(@min,[1 2 3 4 5],[5 4 3 2 1]), [1 2 3 2 1]) >> > assert(map(@min,rand(1,5),[0 0 0 0 0]), [0 0 0 0 0]) >> > assert(map(@(x,y) (sin(x).^2 + >> > cos(y).^2),-pi:0.5:+pi,-pi:0.5:+pi),ones(1,13)) >> > !!!!! test failed >> > assert (map (@(x, y) (sin (x) .^ 2 + cos (y) .^ 2); >> > , -pi:0.5:+pi, -pi:0.5:+pi),ones (1, 13)) expected >> > 1 1 1 1 1 1 1 1 1 1 1 1 1 >> > but got >> > 1 1 1 1 1 1 1 1 1 1 1 1 1 >> > >> > >> > The problem is that Octave 3.2 isn't consistent in what ones() returns: >> > ... >> > GNU Octave, version 3.2.4 >> > ... >> > octave:1> typeinfo(ones(1,13)) >> > ans = range >> > octave:2> typeinfo(ones(2,13)) >> > ans = matrix >> > >> > I don't have the build power on my notebook to build 3.4. Can you check >> > wether >> > this has changed there? >> > >> >> octave:1> typeinfo(ones(1,2)) >> ans = range >> octave:2> typeinfo(ones(2,2)) >> ans = matrix >> octave:3> version >> ans = 3.5.0+ >> >> But I don't think this is a problem; it's a useful optimisation. > > I have a bit of a problem accepting > > octave:3> assert([1,1,1], ones(1,3)) > error: assert ([1, 1, 1],ones (1, 3)) expected > 1 1 1 > but got > 1 1 1 > Type matrix != range > > as not being a bug. This has nothing to do with map().
Oh, I'm sorry, as I said, the assert doesn't fail if you replace map with arrayfun in the current dev sources possibly also in 3.4.0, in particular, your example doesn't fail either anymore. I suppose that has been fixed, then. I have a feeling Octave is nearing a 3.4.1 release, just asked the maintainers list if this is the case. In that case, when we package that, this bug will go away? - Jordi G. H. -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
