Thread Subject: A simple question about FOR

Dear Colleagues,

I am wondering, if there was a simple way to suppress 'for' command in
MATLAB.
This would force students to use smarter solutions than just a lot of
for loops and waiting :)
If You have some hint how to do it in ver.2007b, let me know.
Best Regards
Kamil

Kamil <milka-wywal@to.free.of.pl> wrote in message
<fq1lcs$t07$1@atlantis.news.tpi.pl>...
> Dear Colleagues,
>
> I am wondering, if there was a simple way to
suppress 'for' command in
> MATLAB.
> This would force students to use smarter solutions than
just a lot of
> for loops and waiting :)
> If You have some hint how to do it in ver.2007b, let me
know.
> Best Regards
> Kamil

class policy??? -10 points for every 'for' in an
assignment.

In article <fq1m0u$qlo$1@fred.mathworks.com>, dave@bigcompany.com
says...
> Kamil <milka-wywal@to.free.of.pl> wrote in message
> <fq1lcs$t07$1@atlantis.news.tpi.pl>...
> > Dear Colleagues,
> >
> > I am wondering, if there was a simple way to
> suppress 'for' command in
> > MATLAB.
> > This would force students to use smarter solutions than
> just a lot of
> > for loops and waiting :)
> > If You have some hint how to do it in ver.2007b, let me
> know.
> > Best Regards
> > Kamil
>
> class policy??? -10 points for every 'for' in an
> assignment.
>

My thesis advisor used to deduct 5 points :-)

--
Loren
http://blogs.mathworks.com/loren/

In article <fq1lcs$t07$1@atlantis.news.tpi.pl>,
Kamil <milka-wywal@to.free.of.pl> wrote:

>I am wondering, if there was a simple way to suppress 'for' command in
>MATLAB.
>This would force students to use smarter solutions than just a lot of
>for loops and waiting :)

Sometimes using a 'for' or 'while' loop *is* the most efficient way.

For example a couple of weeks ago, I had a case in which the
standard matlab method of writing a section of code was roughly

[X,Y] = ndgrid(A,B);
answer = sum(X < Y);

Now allow A and B to grow to roughly 1200 elements each, and
call this code 100,000 times, and these two lines of code took
approximately 26.13 minutes to execute.

After some thought, I replaced this short code with a 'while'
loop that took noticably more coding lines, but was linear in the
size of A (and independant of the size of B). With the same dataset,
calling the same 100,000 times, the replacement code took 2.39 seconds
total to execute.

Now, if I were to turn in that 'while' solution that was ~50 times
faster than anyone else achieved using matlab vectorized operations,
would you fail me on the assignment for not using a sufficiently
"smart" solution?
--
  "The beauties of conception are always superior to those of
   expression." -- Walter J. Phillips