«

Does anyone agree with me?
If you have used Matlab, welcome to discuss it.

Allen wrote:
> Does anyone agree with me?
> If you have used Matlab, welcome to discuss it.

Sure, and earth is a heavy version of a basketball. If all you have is
a hammer...

George

> Sure, and earth is a heavy version of a basketball. If all you have is
> a hammer...
>

It is not make sense to compare earth and basketball.
I think Python introduced many idea of matlab.

If you have used matlab, you will say that they are very very similar,
except that matlab was born years earlier and is used mainly in the
area
of matrix calculation.

I do not mean Python shall feel ashamed for it. We will be pleased that
Python
does absorb many successful ideas of computer languages.

Allen wrote:

> It is not make sense to compare earth and basketball.

why not? they're both round, so surely they must have been inspired
by each other. the question is if humanity invented balls before we
figured out that the earth is round, or if it's the other way around...

</F>

On 8 dic, 06:57, "Allen" <che...@naritech.cn> wrote:
>
> If you have used matlab, you will say that they are very very similar,
> except that matlab was born years earlier and is used mainly in the
> area
> of matrix calculation.
>
> I do not mean Python shall feel ashamed for it. We will be pleased that
> Python
> does absorb many successful ideas of computer languages.

Personally, I dont like the Matlab language. I like the "no surprises"
approach of Python, the consistency along different objects, and of
course the vast builtin data structures that let me model almost
anything. None of these are present in Matlab.
I hate gotchas like the degenerate case of sum(M) for a one-row matrix
(avoidable, yes, but "Special cases aren't special enough to break the
rules.").
Of course, if all you do is working with matrices, it's wonderful!

--
Gabriel Genellina

Fredrik Lundh <> writes:

> Allen wrote:
>
> > It is not make sense to compare earth and basketball.
>
> why not? they're both round, so surely they must have been inspired
> by each other. the question is if humanity invented balls before we
> figured out that the earth is round, or if it's the other way around...
>
> </F>
>

In keeping with the computer science flavor, we can say that no one
invented earth or balls to be round. Rouindness is an emergent
behavior of a substance which has shape-forming adhesion and
shape-changing fluidity, and is subject to radially symmetric
shape-impacting processes. Magma and gravity for the earth, leather
and air pressure for inflated balls, sand and accretion for beach
"cannonballs", and snow and hand pressure for snowballs.

--
Harry George
PLM Engineering Architecture

On Dec 7, 11:48 pm, "Allen" <che...@naritech.cn> wrote:
> Does anyone agree with me?
> If you have used Matlab, welcome to discuss it.

Numpy definitely was inspired in its extended array syntax by matlab.
Besides that, I don't think two languages could be more different.
Philosophically, matlab is closer to perl.

-Mike

Allen wrote:
> Does anyone agree with me?
> If you have used Matlab, welcome to discuss it.

Matlab is a tool for scientists and engineers.
Python is a tool for programmers.

I think you are looking at Python from the scientist perspective.
Python's numpy and matplotlib modules would probably feel familiar to
anyone with some matlab experience. But these are not a part of the
language - it is not even a part of the standard library.

I will not go deep into the programmer perspective.
Some more programmer tools: Java, Tcl, Perl, PHP, Lisp, Ruby. Comparing
Python to these makes sense. I think comparing Matlab to any of those
would be absurd. Have a look at modpython.org - is there a need for a
similar modmatlab?

Now, back to the scientist perspective.

In 1999, I was starting my M.Sc. in astrophysics and had to select my
data analysis tools. I needed the standard scientific tools:
scripting, numerics, graphics - Matlab + shell is good enough for
this. But I also needed a library for FITS file processing, which was
not available in Matlab. So Matlab alone was not enough.

Matlab + IRAF + shell was one alternative. Shell + IDL (Interactive
Data Language) was another. There were also other possibilities
(Fortran for numerics, C or Ftools for FITS). To cut it short, after a
while I ended up with shell + IDL as my main tools, occasionally using
the others.

About two years later my scripts were so complex I decided to learn a
scripting language. I was lucky enough to choose Python. Soon I found
pyraf, pyfits and numarray, later gnuPlot.py and matplotlib - IDL was
no longer needed. Python was enough.

Then one day I was looking for a postdoc position. I found something
else, and now I do text mining. I still need the science tools:
scripting, numerics, graphics.

I also need:
1) Regular expressions
2) XML library
3) Database interface

Python covers it all. I think Matlab has a Database interface, but how
about the others?

--
Juho Schultz

Allen wrote:

> Does anyone agree with me?
> If you have used Matlab, welcome to discuss it.

I'm sorry Allen, but Python is heading on the long road to being Lisp.
Matlab will have to wait its turn