MATLAB Central Newsreader - Review of the new Symbolic toolbox included with MATLAB 2008b

Review of the new Symbolic toolbox included with MATLAB 2008b

michael.p.croucher@googlemail.com — Fri, 31 Oct 2008 12:08:35 -0400

Hi All

I have written a review of the new Symbolic toolbox for MATLAB 2008b
over at my blog and thought that some of you on here might be
interested.

http://www.walkingrandomly.com/?p=178

Best Wishes,
Mike

Review of the new Symbolic toolbox included with MATLAB 2008b

Husam Aldahiyat — Sun, 02 Nov 2008 04:14:01 -0500

Nice read. And funny too.

Did atan and atan2 change in this new release? Can you please check for me, particularly atan2's handling of imaginary parts?

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

michael.p.croucher@googlemail.com — Mon, 03 Nov 2008 10:44:55 -0500

Hi Vladimir

> I spotted that you mentioned not a single howler or even bug
> in MATLAB 2008b?

Well I did mention a couple of relatively straightforward integrals
that MATLAB 2008b couldn't do. So far i haven't found any other
problems in MATLAB 2008b but as soon as I do I will write about
them.

> Was is your intention not to put any MATLAB's Symbolic Toolbox
> howler in your Review? Just curious.

Quite the opposite. If I find any bugs in any software system I
support I am usually quite vociferous about it as the support teams of
several large software companies will testify. I simply haven't found
any in 2008b yet - I don't have the benefit of your automated testing
system. Rest assured that, as soon as I do, I will write about them
on my blog, report them to Mathworks and see if they exist in any
other packages.

Best Wishes,
Mike

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

michael.p.croucher@googlemail.com — Mon, 03 Nov 2008 10:46:05 -0500

Husam,

No problem at all. If you give me some sample code to evaluate then I
will send you the comparison.

Cheers,
Mike

On Nov 2, 4:14=A0am, "Husam Aldahiyat" <numand...@gmail.com> wrote:
> Nice read. And funny too.
>
> Did atan and atan2 change in this new release? Can you please check for m=
e, particularly atan2's handling of imaginary parts?

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

Husam Aldahiyat — Mon, 03 Nov 2008 14:46:01 -0500

There's no need. Thanks.

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

Scott Seidman — Fri, 21 Nov 2008 23:09:27 -0500

Vladimir Bondarenko <vb@cybertester.com> wrote in news:c57290e0-7a52-4653-
b0ad-7ef8e2302d13@33g2000yqm.googlegroups.com:

> Greeting to our NASA visitors!
>
> We'd just like to make it clear that now we are in the
> process of lifting the restrictions we had at the early
> stage of the VM machine development.
>
> It means that the new modules we are adding will result
> in higher performance, more subtle automated defects
> classification, - and more.
>
> Best wishes,
>
> Vladimir Bondarenko
>
> Co-founder, CEO, Mathematical Director
>
> http://www.cybertester.com/ Cyber Tester Ltd.
>

Cool. You don't see many textbook examples of someone actively trying to
get promoted from common spammer to wannabe net loon.

--
Scott
Reverse name to reply

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

John D'Errico — Sat, 22 Nov 2008 11:25:04 -0500

Vladimir Bondarenko <vb@cybertester.com> wrote in message <31460dd1-8b41-42d7-ba09-3fcad4da3ed6@w22g2000yqd.googlegroups.com>...
> On Nov 22, 1:09=A0am, Scott Seidman <namdiestt...@mindspring.com> wrote:

> It is my sheer pleasure to introduce to the community glorious
> Dr. Scott H. Seidman (with Fortran roots, as he himself states)
> from University of Rochester.

This is inappropriate and unprofessional, as are many
of your posts. I too will start to flag your stuff as spam.

John

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

Vladimir Bondarenko — Sat, 22 Nov 2008 13:52:47 -0500

On Nov 22, 1:25=A0pm, "John D'Errico" <woodch...@rochester.rr.com>
wrote:
> Vladimir Bondarenko <v...@cybertester.com> wrote in message <31460dd1-8b4=
1-42d7-ba09-3fcad4da3...@w22g2000yqd.googlegroups.com>...
> > On Nov 22, 1:09=3DA0am, Scott Seidman <namdiestt...@mindspring.com> wro=
te:
> > It is my sheer pleasure to introduce to the community glorious
> > Dr. Scott H. Seidman (with Fortran roots, as he himself states)
> > from University of Rochester.
>
> This is inappropriate and unprofessional, as are many
> of your posts. I too will start to flag your stuff as spam.
>
> John

"I too will start to flag your stuff as spam."

I have the only question.

What you are going to do with the defects we calculate in MATLAB?

Flag them, too?

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

John D'Errico — Sat, 22 Nov 2008 15:32:01 -0500

Vladimir Bondarenko <vb@cybertester.com> wrote in message <5d6d0768-5af0-427a-bc81-9538f2b70d10@41g2000yqf.googlegroups.com>...
> On Nov 22, 1:25=A0pm, "John D'Errico" <woodch...@rochester.rr.com>
> wrote:
> > Vladimir Bondarenko <v...@cybertester.com> wrote in message <31460dd1-8b4=
> 1-42d7-ba09-3fcad4da3...@w22g2000yqd.googlegroups.com>...
> > > On Nov 22, 1:09=3DA0am, Scott Seidman <namdiestt...@mindspring.com> wro=
> te:
> > > It is my sheer pleasure to introduce to the community glorious
> > > Dr. Scott H. Seidman (with Fortran roots, as he himself states)
> > > from University of Rochester.
> >
> > This is inappropriate and unprofessional, as are many
> > of your posts. I too will start to flag your stuff as spam.
> >
> > John
>
> "I too will start to flag your stuff as spam."
>
> I have the only question.
>
> What you are going to do with the defects we calculate in MATLAB?
>
> Flag them, too?

If you find a bug, then send it in as a bug report.
There is a place for those reports. Use it. Self
serving self promotion as you have done it is
spam however, and many others have been
recognizing it as such.

This treatment of Scott Seidman is unprofessional
and unacceptable however. You have only succeeded
in making yourself look bad by this post.

John

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

Scott Seidman — Sun, 23 Nov 2008 17:53:26 -0500

Vladimir Bondarenko <vb@cybertester.com> wrote in news:31460dd1-8b41-42d7-
ba09-3fcad4da3ed6@w22g2000yqd.googlegroups.com:

> It is my sheer pleasure to introduce to the community glorious
> Dr. Scott H. Seidman (with Fortran roots, as he himself states)
> from University of Rochester.

Once again, I stand corrected. Obviously, he's trying for net loon and
cyber stalker all at the same time.

We're all thrilled that you can look up shortcomings in Matlabs symbolic
engine, test them in Matlab, and find that they're still shortcomings.

It's a wonderful business strategy to show the world that you're a loon,
and hope that there are still people willing to do business with you.

--
Scott
Reverse name to reply

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

Kenneth Eaton — Sun, 23 Nov 2008 21:02:01 -0500

Vladimir Bondarenko <vb@cybertester.com> wrote in message <31460dd1-8b41-42d7-ba09-3fcad4da3ed6@w22g2000yqd.googlegroups.com>...
>
> It is my sheer pleasure to introduce to the community glorious
> Dr. Scott H. Seidman (with Fortran roots, as he himself states)
> from University of Rochester.

Pardon me, Vlado, but it is an exceedingly ignorant and foolhardy strategy to try and peddle your wares by insulting potential customers in clear view of other potential customers. Making condescending remarks about the research of a learned individual (in this case a professor!) is likely to cause more harm to your credibility than you realize.

Ken

Re: Review of the new Symbolic toolbox included with MATLAB 2008b

Dan Haeg — Mon, 24 Nov 2008 00:37:01 -0500

Vladimir Bondarenko <vb@cybertester.com> wrote in message
> Indeed, having bought MATLAB 2008b for $4137.50 I should have
> guessed that it was very stupid of my to ask MATLAB to integrate
> 1/(1+exp(z)) in a direct way
>
> MATLAB 2008b>> int(1/(1+exp(z)), z,0,inf)
> ans =3D
> 0
>
> Instead, had I had any little gray cell, I should have used
>
> MATLAB 2008b>> f=3Dint(1/(1+exp(z)),z);limit(f,z,inf)-limit(f,z,0)
> ans =3D
> log(2)
>
> Yeah, this is some silly paperwork and waste of time, so what?

MATLAB 2008a
>> syms z
>> int(1/(1+exp(z)), z,0,inf)
ans =
log(2)