Re: problem with consistency of root finding and sorting

[Ben Abbott]

On Wednesday, October 10, 2007, at 01:20PM, "Henry F. Mollet" <address@hidden> 
wrote:
>Ben,
>I fixed the problem with residue.m at my end. Then I used mpoles.m that was
>attached to a recent posting of yours but had no luck with it.
>error: can't perform indexing operations for <unknown type> type
> (see below)
>Henry
>octave-2.9.14:1> b = [1 0 1];
>octave-2.9.14:2> a = [1 0 18 0 81];
>octave-2.9.14:3> [r, p, k, e] = residue(b, a)r =
>
>  -7.2906e-27 - 9.2593e-02i
>  2.2222e-01 - 9.9781e-10i
>  -1.4023e-25 + 9.2593e-02i
>  2.2222e-01 + 9.9781e-10i
>
>p =
>
>  -0.0000 + 3.0000i
>   0.0000 + 3.0000i
>  -0.0000 - 3.0000i
>   0.0000 - 3.0000i
>
>k = [](0x0)
>e =
>
>   1
>   2
>   1
>   2
>
>octave-2.9.14:4> pwd
>ans = /Users/hfm
>octave-2.9.14:5> cd documents/cando2
>octave-2.9.14:6> ls
>OctaveBUP&Misc          mpoles.m                surfaceplot.patch
>gcd2.m                  octave-core             test.m
>jpgread.oct             octave2.m               test_mesh_2_1_71.m
>jpgwrite.oct            residue4.m              test_mesh_2_9_14.m
>octave-2.9.14:7> [e, indx] = mpoles (p)
>error: can't perform indexing operations for <unknown type> type
>octave-2.9.14:7> 
>
>***********************************************************
>on 10/9/07 7:11 PM, Ben Abbott at address@hidden wrote:
>
>> Henry,
>> 
>> Try the one checked into CVS. You can get it here.
>> 
>> http://velveeta.che.wisc.edu/cgi-bin/cvsweb.cgi/octave/scripts/
>> polynomial/
>> 
>> You'll also need to mpoles.m for it to function correctly.
>> 
>> 
>

Henry, 

Since there has been some problem with getting my version of residue to work on 
all platforms, I'm still on the look out for the unfortunate possibility of 
more bugs ... arrgghhh :-(

If by chance you are using cvs to keep everthing as bleeding-edge as possible, 
please do an update and build.

If not, please grab copies of residue.m and mpoles.m from the web interface to 
Octave's cvs server.

http://velveeta.che.wisc.edu/cgi-bin/cvsweb.cgi/octave/scripts/polynomial/

At present, I believe the versions checked in are functioning correctly.

If you place these to files into a folder (by themselves and without renaming 
them) and run octave from there, you'll be able to verify all is working by 
issuing the command "test residue"

I'd love to hear that everything works on your end, so that I can move on to 
other improvements/modifications.

Ben