Thread Subject: Matlab vs. VBA

I use VBA to develop applications for Excel. Came across Matlab
as another high level language. How does it compare with VBA
and can you convert each other's codes? Thanks for your advice.

Tom

"Tom" <tclim36@gmail.com> wrote in message <w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com>...
> I use VBA to develop applications for Excel. Came across Matlab
> as another high level language. How does it compare with VBA
> and can you convert each other's codes? Thanks for your advice.
>
> Tom

You can do what you want in any language. Of course,
your ability and skills in the language you use may limit
your ability to do what you want.

How does MATLAB compare to VBA? A silly question. Of
course it is better than VBA, since otherwise all of us
MATLAB users would be using VBA. A silly question
deserves a silly answer. If you really want a useful
answer, you might tell us what you are doing, and what
you want to do.

Finally, can you convert one language to another? Sure.
Just rewrite the code. This clearly applies to any pair of
languages. It does presume that you know both languages.

John

"Tom" <tclim36@gmail.com> wrote in message
news:w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com...
> I use VBA to develop applications for Excel. Came across Matlab
> as another high level language. How does it compare with VBA
> and can you convert each other's codes? Thanks for your advice.

I don't know of a point-by-point comparison between MATLAB and VB; perhaps
if you share a bit more information about the applications you have written
and are considering converting someone can offer suggestions about how easy
a conversion it will be (or if there's a function directly in MATLAB or one
of the toolboxes to do what you want.)

--
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ
To contact Technical Support use the Contact Us link on
http://www.mathworks.com

Tom wrote:
> I use VBA to develop applications for Excel. Came across Matlab
> as another high level language. How does it compare with VBA
> and can you convert each other's codes? Thanks for your advice.

The biggest difference between VBA/Excel and Matlab is that Matlab is
_NOT_ a spreadsheet paradigm. The second biggest is that Matlab (short
for MATrix LABoratory) is a vectorized language instead of looping
construct as is VB(A).

Those are only two of the most fundamental differences--I'd say it's an
apples/oranges comparison the paradigms are so dissimilar in style and
structure.

--

"John D'Errico" <woodchips@rochester.rr.com> wrote in message
news:i5b7ub$df2$1@fred.mathworks.com...
> "Tom" <tclim36@gmail.com> wrote in message
> <w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com>...
>> I use VBA to develop applications for Excel. Came across Matlab
>> as another high level language. How does it compare with VBA
>> and can you convert each other's codes? Thanks for your advice.
>>
>> Tom
>
> You can do what you want in any language. Of course,
> your ability and skills in the language you use may limit
> your ability to do what you want.
>
> How does MATLAB compare to VBA? A silly question. Of
> course it is better than VBA, since otherwise all of us
> MATLAB users would be using VBA. A silly question
> deserves a silly answer. If you really want a useful
> answer, you might tell us what you are doing, and what
> you want to do.

My question was meant for those who are familiar with both languages
and are able to provide an objective discourse on their differences
e.g. that from dpb. It doesn't call for hubris or a display of arrogance.

> Finally, can you convert one language to another? Sure.
> Just rewrite the code. This clearly applies to any pair of
> languages. It does presume that you know both languages.

There are programs or applications with built-in conversion
capabilities without requiring the user to know both of them.

>
> John

Your comments are lucid and straight to the point.
Thank you for your time.

"dpb" <none@non.net> wrote in message
news:i5ceme$2hd$1@news.eternal-september.org...
> Tom wrote:
>> I use VBA to develop applications for Excel. Came across Matlab
>> as another high level language. How does it compare with VBA
>> and can you convert each other's codes? Thanks for your advice.
>
> The biggest difference between VBA/Excel and Matlab is that Matlab is
> _NOT_ a spreadsheet paradigm. The second biggest is that Matlab (short
> for MATrix LABoratory) is a vectorized language instead of looping
> construct as is VB(A).
>
> Those are only two of the most fundamental differences--I'd say it's an
> apples/oranges comparison the paradigms are so dissimilar in style and
> structure.
>
> --

"Tom" <tclim36@gmail.com> wrote in message <gdseo.4807$Yv.3086@viwinnwfe01.internal.bigpond.com>...
> "John D'Errico" <woodchips@rochester.rr.com> wrote in message
> news:i5b7ub$df2$1@fred.mathworks.com...
> > "Tom" <tclim36@gmail.com> wrote in message
> > <w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com>...
> >> I use VBA to develop applications for Excel. Came across Matlab
> >> as another high level language. How does it compare with VBA
> >> and can you convert each other's codes? Thanks for your advice.
> >>
> >> Tom
> >
> > You can do what you want in any language. Of course,
> > your ability and skills in the language you use may limit
> > your ability to do what you want.
> >
> > How does MATLAB compare to VBA? A silly question. Of
> > course it is better than VBA, since otherwise all of us
> > MATLAB users would be using VBA. A silly question
> > deserves a silly answer. If you really want a useful
> > answer, you might tell us what you are doing, and what
> > you want to do.
>
> My question was meant for those who are familiar with both languages
> and are able to provide an objective discourse on their differences
> e.g. that from dpb. It doesn't call for hubris or a display of arrogance.

You have not stated what you are doing.

Both tools are tremendously flexible. Without that
information, your question is meaningless.


> There are programs or applications with built-in conversion
> capabilities without requiring the user to know both of them.

And almost always, those tools produce terrible code.
Inefficient, difficult to read. Useless.

John

"John D'Errico" <woodchips@rochester.rr.com> wrote in message
news:i5dkkp$aco$1@fred.mathworks.com...
> "Tom" <tclim36@gmail.com> wrote in message
> <gdseo.4807$Yv.3086@viwinnwfe01.internal.bigpond.com>...
>> "John D'Errico" <woodchips@rochester.rr.com> wrote in message
>> news:i5b7ub$df2$1@fred.mathworks.com...
>> > "Tom" <tclim36@gmail.com> wrote in message
>> > <w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com>...
>> >> I use VBA to develop applications for Excel. Came across Matlab
>> >> as another high level language. How does it compare with VBA
>> >> and can you convert each other's codes? Thanks for your advice.
>> >>
>> >> Tom
>> >
>> > You can do what you want in any language. Of course,
>> > your ability and skills in the language you use may limit
>> > your ability to do what you want.
>> >
>> > How does MATLAB compare to VBA? A silly question. Of
>> > course it is better than VBA, since otherwise all of us
>> > MATLAB users would be using VBA. A silly question
>> > deserves a silly answer. If you really want a useful
>> > answer, you might tell us what you are doing, and what
>> > you want to do.
>>
>> My question was meant for those who are familiar with both languages
>> and are able to provide an objective discourse on their differences
>> e.g. that from dpb. It doesn't call for hubris or a display of arrogance.
>
> You have not stated what you are doing.

Please refer to my reply to Steven_Lord.

>
> Both tools are tremendously flexible. Without that
> information, your question is meaningless.
>
>
>> There are programs or applications with built-in conversion
>> capabilities without requiring the user to know both of them.
>
> And almost always, those tools produce terrible code.
> Inefficient, difficult to read. Useless.
>
> John

On 29/08/10 8:55 AM, Tom wrote:

> Please refer to my reply to Steven_Lord.

I would, if I could *find* a reply from you to Steve Lord. I _have_ been
able to find a reply from you to dpb, but unfortunately that reply said
nothing about what you were doing.

 > My question was meant for those who are familiar with both languages
 > and are able to provide an objective discourse on their differences

You are aware that such things are what _multiple_ books get written
about??

"Walter Roberson" <roberson@hushmail.com> wrote in message
news:XAueo.31525$co1.891@newsfe11.iad...
> On 29/08/10 8:55 AM, Tom wrote:
>
>> Please refer to my reply to Steven_Lord.
>
> I would, if I could *find* a reply from you to Steve Lord.

I wonder why you are unable to *find* that?
It is clearly visible on my screen. There is a
brief explanation plus an attachment from
which more information can be gleaned.

> I _have_ been > able to find a reply from you to dpb, but unfortunately
> that reply said nothing about what
> you were doing.

Ok. If I ask for a comparison of OS like Microsoft Windows
and GNU/Linux, I don't expect to be asked what I was going
to do with them. To do so would limit the scope of their
comparative differences to what I was going to use them for.
I was expecting to get a rounded and objective dissertation.
The answer from "dpb" did just that. But one respondent
took a narrow and gladiatorial focus which says more about
himself than the legitimacy of my question.

> > My question was meant for those who are familiar with both languages
> > and are able to provide an objective discourse on their differences
>
> You are aware that such things are what _multiple_ books get written
> about??

Tom wrote:
> "Walter Roberson" <roberson@hushmail.com> wrote in message
> news:XAueo.31525$co1.891@newsfe11.iad...
>> On 29/08/10 8:55 AM, Tom wrote:
>>
>>> Please refer to my reply to Steven_Lord.
>>
>> I would, if I could *find* a reply from you to Steve Lord.
>
> I wonder why you are unable to *find* that?
> It is clearly visible on my screen. There is a
> brief explanation plus an attachment from
> which more information can be gleaned.
...

Didn't get propagated to the eternal-september usenet mirror, either...

--

On Aug 29, 3:51 am, dpb <n...@non.net> wrote:
> Tom wrote:
> > I use VBA to develop applications for Excel. Came across Matlab
> > as another high level language. How does it compare with VBA
> > and can you convert each other's codes? Thanks for your advice.
>
> The biggest difference between VBA/Excel and Matlab is that Matlab is
> _NOT_ a spreadsheet paradigm.  

Agreed.

> The second biggest is that Matlab (short
> for MATrix LABoratory) is a vectorized language instead of looping
> construct as is VB(A).

First half correct; second half wrong.

Yes, matlab started out as the MATrix LABoratory; no, there is no
such
thing as a 'vectorized language'. Any talk about 'vectorization' can
be traced back to a bug / design flaw of matlab's that took a couple
of decades to find and correct. In the mean time, matlab users worked
around the bug by developing the coding style that is unique to
matlab
and known as 'vectorizing'.

> Those are only two of the most fundamental differences--I'd say it's an
> apples/oranges comparison the paradigms are so dissimilar in style and
> structure.

Don't know VBA, but matlab is by now just about the only remaining
ghost
of the useful tools of the olden days: Command line interface,
scripting,
programming, some rather useful libraries and toolboxes. I prefer
matlab
because it's a Jack of all trades (if a master of none) that allow me
to
work in the more productive ways. No hazzle with GUIs, no
workarounds,
interface with the efficient tools when you need to.

I think you might find that matlab allows the user who understands
computers to be far more efficient than he would be with VBA.

Rune

Rune Allnor wrote:
> On Aug 29, 3:51 am, dpb <n...@non.net> wrote:
>> Tom wrote:
>>> I use VBA to develop applications for Excel. Came across Matlab
>>> as another high level language. How does it compare with VBA
>>> and can you convert each other's codes? Thanks for your advice.
>> The biggest difference between VBA/Excel and Matlab is that Matlab is
>> _NOT_ a spreadsheet paradigm.
>
> Agreed.
>
>> The second biggest is that Matlab (short
>> for MATrix LABoratory) is a vectorized language instead of looping
>> construct as is VB(A).
>
> First half correct; second half wrong.
>
> Yes, matlab started out as the MATrix LABoratory; no, there is no
> such thing as a 'vectorized language'. ...

Depends on what you mean by a "vectorized language". There certainly is
by what I mean by it -- you can devise your own nomenclature if you wish.

--

On Aug 30, 6:15 am, dpb <n...@non.net> wrote:
> Rune Allnor wrote:
> > On Aug 29, 3:51 am, dpb <n...@non.net> wrote:
> >> Tom wrote:
> >>> I use VBA to develop applications for Excel. Came across Matlab
> >>> as another high level language. How does it compare with VBA
> >>> and can you convert each other's codes? Thanks for your advice.
> >> The biggest difference between VBA/Excel and Matlab is that Matlab is
> >> _NOT_ a spreadsheet paradigm.  
>
> > Agreed.
>
> >> The second biggest is that Matlab (short
> >> for MATrix LABoratory) is a vectorized language instead of looping
> >> construct as is VB(A).
>
> > First half correct; second half wrong.
>
> > Yes, matlab started out as the MATrix LABoratory; no, there is no
> > such thing as a 'vectorized language'.  ...
>
> Depends on what you mean by a "vectorized language".  There certainly is
> by what I mean by it -- you can devise your own nomenclature if you wish.

I know this is controversial, but I prefer to stick with what
is already in common use. The fact of the matter is that no one
that is not familiar with matlab will not have the faintest clue
what 'vectoized code' might mean. The reason is simple: The term
'vectorized' is only used in the matlab community. And the reason
for *that* is Matlab's infamous interpreter bug / design flaw.

Rune

On Aug 30, 4:34 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
> On Aug 30, 6:15 am, dpb <n...@non.net> wrote:
>
>
>
> > Rune Allnor wrote:
> > > On Aug 29, 3:51 am, dpb <n...@non.net> wrote:
> > >> Tom wrote:
> > >>> I use VBA to develop applications for Excel. Came across Matlab
> > >>> as another high level language. How does it compare with VBA
> > >>> and can you convert each other's codes? Thanks for your advice.
> > >> The biggest difference between VBA/Excel and Matlab is that Matlab is
> > >> _NOT_ a spreadsheet paradigm.  
>
> > > Agreed.
>
> > >> The second biggest is that Matlab (short
> > >> for MATrix LABoratory) is a vectorized language instead of looping
> > >> construct as is VB(A).
>
> > > First half correct; second half wrong.
>
> > > Yes, matlab started out as the MATrix LABoratory; no, there is no
> > > such thing as a 'vectorized language'.  ...
>
> > Depends on what you mean by a "vectorized language".  There certainly is
> > by what I mean by it -- you can devise your own nomenclature if you wish.
>
> I know this is controversial, but I prefer to stick with what
> is already in common use. The fact of the matter is that no one
> that is not familiar with matlab will not have the faintest clue
> what 'vectoized code' might mean. The reason is simple: The term
> 'vectorized' is only used in the matlab community. And the reason
> for *that* is Matlab's infamous interpreter bug / design flaw.
>
> Rune

What utter crap.
The concept of vectorized code/language (and indeed the term itself)
is used all the time when dealing with code for supercomputers.

On Aug 30, 6:45 am, TideMan <mul...@gmail.com> wrote:
> On Aug 30, 4:34 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
>
>
>
>
>
> > On Aug 30, 6:15 am, dpb <n...@non.net> wrote:
>
> > > Rune Allnor wrote:
> > > > On Aug 29, 3:51 am, dpb <n...@non.net> wrote:
> > > >> Tom wrote:
> > > >>> I use VBA to develop applications for Excel. Came across Matlab
> > > >>> as another high level language. How does it compare with VBA
> > > >>> and can you convert each other's codes? Thanks for your advice.
> > > >> The biggest difference between VBA/Excel and Matlab is that Matlab is
> > > >> _NOT_ a spreadsheet paradigm.  
>
> > > > Agreed.
>
> > > >> The second biggest is that Matlab (short
> > > >> for MATrix LABoratory) is a vectorized language instead of looping
> > > >> construct as is VB(A).
>
> > > > First half correct; second half wrong.
>
> > > > Yes, matlab started out as the MATrix LABoratory; no, there is no
> > > > such thing as a 'vectorized language'.  ...
>
> > > Depends on what you mean by a "vectorized language".  There certainly is
> > > by what I mean by it -- you can devise your own nomenclature if you wish.
>
> > I know this is controversial, but I prefer to stick with what
> > is already in common use. The fact of the matter is that no one
> > that is not familiar with matlab will not have the faintest clue
> > what 'vectoized code' might mean. The reason is simple: The term
> > 'vectorized' is only used in the matlab community. And the reason
> > for *that* is Matlab's infamous interpreter bug / design flaw.
>
> > Rune
>
> What utter crap.
> The concept of vectorized code/language (and indeed the term itself)
> is used all the time when dealing with code for supercomputers.

In that context the term applies to instruction scheduling on the
machine code level - that is, lower even than assembly language.
Vectorized architectures tend to start executing the 2nd instruction
before the 1st has finished. It's a completely different thing
than the matlab voodoo.

Rune

On 29/08/10 7:46 PM, Tom wrote:
> "Walter Roberson" <roberson@hushmail.com> wrote in message
> news:XAueo.31525$co1.891@newsfe11.iad...
>> On 29/08/10 8:55 AM, Tom wrote:
>>
>>> Please refer to my reply to Steven_Lord.
>>
>> I would, if I could *find* a reply from you to Steve Lord.
>
> I wonder why you are unable to *find* that?
> It is clearly visible on my screen.

Please post a URL at which we can find a copy of that message.

I have the headers for the last 10000 postings in this resource on my
system, and I read through every message from you (at least under that
name) from that assortment without finding it. I have also checked
Mathwork's copy of the thread, but cannot see it there either.
http://www.mathworks.com/matlabcentral/newsreader/view_thread/290409

 > Ok. If I ask for a comparison of OS like Microsoft Windows
 > and GNU/Linux, I don't expect to be asked what I was going
 > to do with them.

Then your expectations would not match the reality of any of the
electronic forums I have worked over the last 20 years.

Your question is like "Please compare wood-working and metal-working,
and don't make any assumptions about what I might want to do". Well,
there is estimated to be 100,000 species of trees and then there are
50,000 species of monocots, some of which produce wood (e.g., bamboo,
bananas) and others of which do not (e.g., wheat); and there are
different techniques or optimums for working with each -- and different
woods or grasses are better suited for smelting of different metals, but
the smelting temperature to use depends on desired use of the metal
including cooling schedule and whether you are after the mechanical
properties or electrical properties, and ...

Some questions are just too big to answer meaningfully in any realistic
amount of time and effort.

Are you interested in the details of memory management? In the
multiprocessing primitives available? In the implemented automatic
parallelization? In the trade-offs that have been made for efficiency vs
ensuring the environment does not crash vs security vs checking that the
user isn't trying to do something unwise? In the ease of interfacing
with other programming languages, and if so which ones since the other
languages have their own quirks? In the ease of networking
communications? In the multilingual support? Interested in real-time
work and response time guarantees? Real-time and memory control to
prevent application memory overflow? Real-time and in what the semantics
of the language imply about typical and maximum latencies? Interested in
the ease of expressing code in BNF (Brackus-Naur Form) ? In the ease of
expressing the denotational semantics of programs coded in the language?
Interested in comparisons of the orthogonality of features? Interested
in whether the language is structured in such a way that features which
are expensive to support do not have add any overhead to programs that
do not happen to require them? Interested in support for incremental
compilation? In whether it supports Just In Time (JIT) compilation, and
if so at what resolution? Interested in the debugging features
available? Is the language "reflective" such as the programming language
Scheme is? Is the language static execution only, or does it support
dynamic execution of constructed strings or of constructed function
structures, and if it supports dynamic execution what the overhead
involved with that is? Is your primary interest in legal licensing
terms? In whether Public License versions of the software are available?

I could go on, but I have already presented a page-full just of
_categories_ of what someone could potentially be interested in when
they ask to compare two languages.

You know, this whole discussion would have been a lot easier if you had
merely asked something like "What are the traditional strengths of these
languages?", but you didn't ask that: you asked to _compare_ them and
you want that comparison without guiding us as to what interests you and
what does not.

If you want a detailed dissection comparing two programming languages
from multiple viewpoints, you need perfectionist analysts who together
have fairly broad knowledge bases, and you need writers able to express
the complex information in a readable form. To get such a comparison in
an electronic forum from a single person working alone is unlikely --
unlikely that you would find such a person who combines the technical
skills, writing skills, and appropriate knowledge. And if you _could_
find such a person, it is unlikely that they would not ask a question
equivalent to "Well, where do you want the answer to _start_ ?"

> Ok. If I ask for a comparison of OS like Microsoft Windows
> and GNU/Linux, I don't expect to be asked what I was going
> to do with them. To do so would limit the scope of their
> comparative differences to what I was going to use them for.

Well, you should expect to be asked what you're using it for. If a friend of mine were about to buy/build a computer and asked me whether they should get Windows or Linux, the first thing I would say would be (to get both, but seriously): do you plan on gaming? (Gaming isn't the only example, but there are definite benefits of one OS over the other, and it would be silly to not consider them.)

This is even more true with programming languages. You are much less tied down to one programming language, and you should always try to choose the most appropriate language for the task at hand.

I also don't see a post describing your intended use, but you did say that you develop applications for Excel. Obviously, VBA is already thoroughly integrated into Excel. MATLAB has pretty good Excel integration as well, through the actxserver command. You can control absolutely everything in Excel from MATLAB with code similar to VBA. In fact, although I don't know VBA, I can often use the Excel tools to automatically record macros and then copy and paste the code to MATLAB. Few changes are needed. Through the command line interface in MATLAB, you have a better ability to explore the Excel objects and learn how to manipulate them further.

Also in MATLAB, there are lots of included functions that cover a lot of what you're likely to do. If they don't, there are toolboxes and the File Exchange.

But to get better integration of MATLAB with Excel, you would need to purchase some toolboxes. There is a recorded webinar you can watch about using MATLAB with Excel that makes use of the Spreadsheet Link EX toolbox. What you'll see in that webinar is that it is very easy to auto-generate lots of code in MATLAB and to automate a process you would otherwise have done in Excel. The resulting application turned out to be much faster in MATLAB than in Excel in the webinar, but of course your experience may be different depending on your specific application.

So the things you're likely to gain by using MATLAB are speed and additional functionality by having access to MATLAB, toolboxes, and the FEX. The things you're giving up are money and a little time while you're learning.

Rune Allnor wrote:
...

> ...I prefer to stick with what is already in common use. ...

Which would be???

--

"Tom" <tclim36@gmail.com> wrote in message
news:rbDeo.4862$Yv.4798@viwinnwfe01.internal.bigpond.com...
> "Walter Roberson" <roberson@hushmail.com> wrote in message
> news:XAueo.31525$co1.891@newsfe11.iad...
>> On 29/08/10 8:55 AM, Tom wrote:
>>
>>> Please refer to my reply to Steven_Lord.
>>
>> I would, if I could *find* a reply from you to Steve Lord.
>
> I wonder why you are unable to *find* that?
> It is clearly visible on my screen. There is a
> brief explanation plus an attachment from
> which more information can be gleaned.

You sent your reply to me via direct email; that does NOT get posted to the
newsgroup unless you specifically CC the newsgroup or forward it to the
group later on.

In general, if you ask a question on a newsgroup the best practice is to
keep the responses on the newsgroup until or unless someone asks you to
switch to email discussion.

>> I _have_ been > able to find a reply from you to dpb, but unfortunately
>> that reply said nothing about what
>> you were doing.
>
> Ok. If I ask for a comparison of OS like Microsoft Windows
> and GNU/Linux, I don't expect to be asked what I was going
> to do with them.

Really?

Restating your original question in terms of Windows and Linux, you asked:

"I use Windows. Came across Linux as another operating system. How does it
compare with Windows and can you use the same applications?"

[Yeah, I took a little bit of liberty with the wording of that question.
It's in the spirit of your original question, though.]

Answering this question would be MUCH easier if we knew what you were using
Windows to do. For instance, if you were my mother (who's far from the most
computer-savvy person I know -- sorry Mom) then the answer would be "Mom,
just stick with Windows. You can read web pages on it and play your games
just fine." If you were an IT manager asking for advice about which OS to
run on your servers, then the discussion would be much more in-depth.

> To do so would limit the scope of their
> comparative differences to what I was going to use them for.
> I was expecting to get a rounded and objective dissertation.
> The answer from "dpb" did just that. But one respondent
> took a narrow and gladiatorial focus which says more about
> himself than the legitimacy of my question.

We get people on this newsgroup asking very broad, open-ended questions
relatively frequently. Often, they're people demanding "Here's my homework,
do it for me" and I think that's affected how some of the gurus respond to
those type of broad questions.

--
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ
To contact Technical Support use the Contact Us link on
http://www.mathworks.com

On Aug 30, 3:37 pm, dpb <n...@non.net> wrote:
> Rune Allnor wrote:
>
> ...
>
> > ...I prefer to stick with what is already in common use. ...
>
> Which would be???

That outside the matlab community there is no such thing as
a 'vectorized language'.

Rune

On 30/08/10 9:38 AM, Rune Allnor wrote:

> That outside the matlab community there is no such thing as
> a 'vectorized language'.

Rune, if you supply the citations for non-existence, then someone could
edit the Wikipedia entry on the topic,
http://en.wikipedia.org/wiki/Array_programming

Currently that entry says in part:


In computer science, array programming languages (also known as vector
or multidimensional languages) generalize operations on scalars to apply
transparently to vectors, matrices, and higher dimensional arrays. [...]

The canonical examples of array programming languages are APL, J, and
Fortran 90. Others include: A+, Analytica, IDL, K, Q, Mathematica,
MATLAB, MOLSF, NumPy, GNU Octave, PDL, R, S-Lang, SAC, Nial and ZPL.

Category:Array programming languages provides an exhaustive list.

"Steven_Lord" <slord@mathworks.com> wrote in message
news:i5gckh$cu2$1@fred.mathworks.com...
>
>
> "Tom" <tclim36@gmail.com> wrote in message
> news:rbDeo.4862$Yv.4798@viwinnwfe01.internal.bigpond.com...
>> "Walter Roberson" <roberson@hushmail.com> wrote in message
>> news:XAueo.31525$co1.891@newsfe11.iad...
>>> On 29/08/10 8:55 AM, Tom wrote:
>>>
>>>> Please refer to my reply to Steven_Lord.
>>>
>>> I would, if I could *find* a reply from you to Steve Lord.
>>
>> I wonder why you are unable to *find* that?
>> It is clearly visible on my screen. There is a
>> brief explanation plus an attachment from
>> which more information can be gleaned.
>
> You sent your reply to me via direct email; that does NOT get posted to
> the newsgroup unless you specifically CC the newsgroup or forward it to
> the group later on.
>
> In general, if you ask a question on a newsgroup the best practice is to
> keep the responses on the newsgroup until or unless someone asks you to
> switch to email discussion.

My apology - I accidently clicked the wrong button. But as soon as I
realized that,
I corrected and resent the same to this newsgroup. That is why I can see it
on my
screen.

>>> I _have_ been > able to find a reply from you to dpb, but unfortunately
>>> that reply said nothing about what
>>> you were doing.
>>
>> Ok. If I ask for a comparison of OS like Microsoft Windows
>> and GNU/Linux, I don't expect to be asked what I was going
>> to do with them.
>
> Really?

Yes. Be objective. Confine the discussion to the merits and limitations of
the two
products. What I am going to use it for does not contribute to the merits of
the
the product. It only limits the focus. Whereas a broad discussion can reveal
properties and capabilities that we may not be aware of and thus can lead
to new ideas and usage for the product.

> Restating your original question in terms of Windows and Linux, you asked:
>
> "I use Windows. Came across Linux as another operating system. How does
> it compare with Windows and can you use the same applications?"

Why only "the same applications"? - this is again limiting the horizon.

> [Yeah, I took a little bit of liberty with the wording of that question.
> It's in the spirit of your original question, though.]
>
> Answering this question would be MUCH easier if we knew what you were
> using Windows to do.

This is from your angle. But from my angle I like to find out its full
potentials and
possibly discover new uses for it and not be limited to a single use. If
you think
about it is a positive marketing position to take.

> For instance, if you were my mother (who's far from the most
> computer-savvy person I know -- sorry Mom) then the answer would be "Mom,
> just stick with Windows. You can read web pages on it and play your games
> just fine." If you were an IT manager asking for advice about which OS to
> run on your servers, then the discussion would be much more in-depth.
>
>> To do so would limit the scope of their
>> comparative differences to what I was going to use them for.
>> I was expecting to get a rounded and objective dissertation.
>> The answer from "dpb" did just that. But one respondent
>> took a narrow and gladiatorial focus which says more about
>> himself than the legitimacy of my question.
>
> We get people on this newsgroup asking very broad, open-ended questions
> relatively frequently. Often, they're people demanding "Here's my
> homework, do it for me" and I think that's affected how some of the gurus
> respond to those type of broad questions.

Why get hot under the collar if the question is provocative? Just don't
respond.

>
> --
> Steve Lord
> slord@mathworks.com
> comp.soft-sys.matlab (CSSM) FAQ:
> http://matlabwiki.mathworks.com/MATLAB_FAQ
> To contact Technical Support use the Contact Us link on
> http://www.mathworks.com

You have given the most explicit account of the capabilities of matlab in
its integration with Excel
together with its VBA code handling capability. That is a big plus including
its additional functionality
and processing speed. I definitely would be looking into it. Had I just
mentioned a particular intended
use, I would have missed out on all this information as people would just
tailor their response to that
particular use. I understand that the software has links to external
tutorials for new users. Are there other
sources where I can get information to get started? Thank you for your
advice.

"Andy " <myfakeemailaddress@gmail.com> wrote in message
news:i5g86h$lq0$1@fred.mathworks.com...
>> Ok. If I ask for a comparison of OS like Microsoft Windows
>> and GNU/Linux, I don't expect to be asked what I was going
>> to do with them. To do so would limit the scope of their
>> comparative differences to what I was going to use them for.
>
> Well, you should expect to be asked what you're using it for. If a friend
> of mine were about to buy/build a computer and asked me whether they
> should get Windows or Linux, the first thing I would say would be (to get
> both, but seriously): do you plan on gaming? (Gaming isn't the only
> example, but there are definite benefits of one OS over the other, and it
> would be silly to not consider them.)
>
> This is even more true with programming languages. You are much less tied
> down to one programming language, and you should always try to choose the
> most appropriate language for the task at hand.
>
> I also don't see a post describing your intended use, but you did say that
> you develop applications for Excel. Obviously, VBA is already thoroughly
> integrated into Excel. MATLAB has pretty good Excel integration as well,
> through the actxserver command. You can control absolutely everything in
> Excel from MATLAB with code similar to VBA. In fact, although I don't
> know VBA, I can often use the Excel tools to automatically record macros
> and then copy and paste the code to MATLAB. Few changes are needed.
> Through the command line interface in MATLAB, you have a better ability to
> explore the Excel objects and learn how to manipulate them further.
>
> Also in MATLAB, there are lots of included functions that cover a lot of
> what you're likely to do. If they don't, there are toolboxes and the File
> Exchange.
>
> But to get better integration of MATLAB with Excel, you would need to
> purchase some toolboxes. There is a recorded webinar you can watch about
> using MATLAB with Excel that makes use of the Spreadsheet Link EX toolbox.
> What you'll see in that webinar is that it is very easy to auto-generate
> lots of code in MATLAB and to automate a process you would otherwise have
> done in Excel. The resulting application turned out to be much faster in
> MATLAB than in Excel in the webinar, but of course your experience may be
> different depending on your specific application.
> So the things you're likely to gain by using MATLAB are speed and
> additional functionality by having access to MATLAB, toolboxes, and the
> FEX. The things you're giving up are money and a little time while you're
> learning.

On Aug 30, 5:46 pm, Walter Roberson <rober...@hushmail.com> wrote:
> On 30/08/10 9:38 AM, Rune Allnor wrote:
>
> > That outside the matlab community there is no such thing as
> > a 'vectorized language'.
>
> Rune, if you supply the citations for non-existence, then someone could
> edit the Wikipedia entry on the topic,http://en.wikipedia.org/wiki/Array_programming
>
> Currently that entry says in part:

Where, on that page or in your excerpt, does the term 'vectorized
language'
occur?

There is a difference between a precise, meaningful term like
'language that
lends itself to vector / matrix / array semantics' and mere mumbo-
jumbo
like 'vectorized language'.

The difference is that 'vector semantics' refers to the programming
language
being favourable to the human's prefernce to write

y = A*x;

to indicate a matrix-vector produc, instead of a double loop like

for i = 1:N
    y(i) = 0;
    for j = 1:N
       y(i) = y(i) + A(i,j)*x(i);
    end;
end;

The net effect of the vector *semantics* is that

1) The code becomes less error-prone to write, since there is
   less to write and thus less to go wrong.
2) The code becomes easier to read, since the semantics maps
   more directly to the human-world application domain.

Dbd's flaw or blunder was to not realize the difference between
application-specific *semantics* and more fundamental issues.
It makes no sense to start talking about 'vectoization as a
paradigm' on the source code level. One *might* use the term
at a machine code level, but that's a completely different issue,
that depends on hardware architecture, not software semantics.

Rune

On Aug 30, 8:52 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
.> ...
.>
.> Dbd's flaw or blunder was to not realize the difference ...
.>
.> Rune

Is it safe to assume you mean "dpb"?

Dale B. Dalrymple

On Aug 31, 6:08 am, dbd <d...@ieee.org> wrote:
> On Aug 30, 8:52 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
> .> ...
> .>
> .> Dbd's flaw or blunder was to not realize the difference ...
> .>
> .> Rune
>
> Is it safe to assume you mean "dpb"?
>
> Dale B. Dalrymple

It is. Sorry, Dale.

Rune

"Tom" <tclim36@gmail.com> wrote in message
news:FOXeo.5022$FH2.4496@viwinnwfe02.internal.bigpond.com...
> "Steven_Lord" <slord@mathworks.com> wrote in message
> news:i5gckh$cu2$1@fred.mathworks.com...

*snip*

>>> Ok. If I ask for a comparison of OS like Microsoft Windows
>>> and GNU/Linux, I don't expect to be asked what I was going
>>> to do with them.
>>
>> Really?
>
> Yes. Be objective. Confine the discussion to the merits and limitations of
> the two
> products. What I am going to use it for does not contribute to the merits
> of the
> the product. It only limits the focus. Whereas a broad discussion can
> reveal
> properties and capabilities that we may not be aware of and thus can lead
> to new ideas and usage for the product.

Books can and have been written about the merits and limitations of all of
Microsoft Windows, GNU/Linux, VBA, and MATLAB. Ideally, newsgroup postings
should be shorter than a book.

>> Restating your original question in terms of Windows and Linux, you
>> asked:
>>
>> "I use Windows. Came across Linux as another operating system. How does
>> it compare with Windows and can you use the same applications?"
>
> Why only "the same applications"? - this is again limiting the horizon.

Your original question involved "convert each other's codes" -- "the same
applications" was my analogy to that part of your original questions.

>> [Yeah, I took a little bit of liberty with the wording of that question.
>> It's in the spirit of your original question, though.]
>>
>> Answering this question would be MUCH easier if we knew what you were
>> using Windows to do.
>
> This is from your angle. But from my angle I like to find out its full
> potentials and
> possibly discover new uses for it and not be limited to a single use. If
> you think
> about it is a positive marketing position to take.

Okay, if you want to see the "full potentials", I have a couple of
references that you may be interested in reviewing. The first is the MATLAB
and Simulink Based Books page, which includes descriptions of over 1200
books that make use of MATLAB and/or Simulink in various ways.

http://www.mathworks.com/support/books/

The second is the product list:

http://www.mathworks.com/products/product_listing/index.html

Each individual product's page contains a description of that product, demos
and lists of features present in that product, links to the documentation
for the product, and technical literature and stories about how users have
used that product in their work.

Of course, if you were, say, working in the field of computational biology
(just to choose one particular topic at random) then there's a page that is
targeted toward your particular area of interest.

http://www.mathworks.com/computational-biology/

That way you wouldn't have to weed through all the discussion about topics
in which you're not interested.

>> For instance, if you were my mother (who's far from the most
>> computer-savvy person I know -- sorry Mom) then the answer would be "Mom,
>> just stick with Windows. You can read web pages on it and play your
>> games just fine." If you were an IT manager asking for advice about
>> which OS to run on your servers, then the discussion would be much more
>> in-depth.
>>
>>> To do so would limit the scope of their
>>> comparative differences to what I was going to use them for.
>>> I was expecting to get a rounded and objective dissertation.
>>> The answer from "dpb" did just that. But one respondent
>>> took a narrow and gladiatorial focus which says more about
>>> himself than the legitimacy of my question.
>>
>> We get people on this newsgroup asking very broad, open-ended questions
>> relatively frequently. Often, they're people demanding "Here's my
>> homework, do it for me" and I think that's affected how some of the gurus
>> respond to those type of broad questions.
>
> Why get hot under the collar if the question is provocative? Just don't
> respond.

As far as I know, everyone posting to this group (except for the spam bots
trying to sell knock-off shoes or herbal remedies, of course) is human, and
I can think of very few humans who never get annoyed or frustrated and
express that annoyance or frustration.

--
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ
To contact Technical Support use the Contact Us link on
http://www.mathworks.com

> Yes. Be objective. Confine the discussion to the merits and limitations of
> the two
> products. What I am going to use it for does not contribute to the merits of
> the
> the product. It only limits the focus. Whereas a broad discussion can reveal
> properties and capabilities that we may not be aware of and thus can lead
> to new ideas and usage for the product.

You have to limit your focus. MATLAB and VBA are not inherently good or bad. They're both complete programming languages. They have advantages and disadvantages in specific areas. There are literally thousands of different pros and cons for each language that we could list off, but nobody is going to take the time to do that.


>Had I just
>mentioned a particular intended
>use, I would have missed out on all this information as people would just
>tailor their response to that
>particular use.

That's kind of backwards, since I was responding to the only specific uses you did mention: developing applications that work with Excel and converting code between VBA and MATLAB.



> Why only "the same applications"? - this is again limiting the horizon.

I think this was just an analogy or an example. If somebody were asking me to recommend an OS, the most important thing I'd ask them is what applications they would want to run. That's a major way to determine the correct OS for the user. There are other questions that are, essentially, proxies for this (such as my previous comment about gaming). Certainly there are other pros/cons to consider, but you basically shouldn't consider most of them. It's not worth your time to learn both OSs in sufficient detail to do a full comparison.

I would argue this is more true for programming languages for multiple reasons. As I said before, you are much less tied down to one programming language than you are to one OS. In addition, since both VBA and MATLAB are complete languages, you can do whatever you want in either, unlike with the OS analogy (where some applications really won't run on one OS, making that an important distinguishing characteristic).


> This is from your angle. But from my angle I like to find out its full
> potentials and
> possibly discover new uses for it and not be limited to a single use.

Again, these are complete programming languages. Trying to discover new uses is the wrong approach, since each language can do anything you want it to. You are more likely to care about things like efficiency of workflow, code, and integration with other applications.


So, again, it would really help (both you and us) if you would mention what your intended uses are. These don't need to be specific. (My previous reply was only to the general need for integration with Excel.) But even answering the following questions would be useful:

1) Will you continue storing your data in Excel?
2) Are you looking to continue integrating heavily with Excel? Does Excel need to be the front end for your application? If not, does it need to be part of the application at all?
3) What general heading would you put your application under? (Image processing? Signal processing? Stock analysis? Etc.)
4) Do you need to be able to distribute your application?

There are lots of other important questions, but these will be a good start.

"Steven_Lord" <slord@mathworks.com> wrote in message
news:i5j2e7$ht3$1@fred.mathworks.com...
>
>
> "Tom" <tclim36@gmail.com> wrote in message
> news:FOXeo.5022$FH2.4496@viwinnwfe02.internal.bigpond.com...
>> "Steven_Lord" <slord@mathworks.com> wrote in message
>> news:i5gckh$cu2$1@fred.mathworks.com...
>
> *snip*
>
>>>> Ok. If I ask for a comparison of OS like Microsoft Windows
>>>> and GNU/Linux, I don't expect to be asked what I was going
>>>> to do with them.
>>>
>>> Really?
>>
>> Yes. Be objective. Confine the discussion to the merits and limitations
>> of the two
>> products. What I am going to use it for does not contribute to the merits
>> of the
>> the product. It only limits the focus. Whereas a broad discussion can
>> reveal
>> properties and capabilities that we may not be aware of and thus can
>> lead
>> to new ideas and usage for the product.
>
> Books can and have been written about the merits and limitations of all of
> Microsoft Windows, GNU/Linux, VBA, and MATLAB. Ideally, newsgroup
> postings should be shorter than a book.

But statistically limited to a handful of helpful people.
>
>>> Restating your original question in terms of Windows and Linux, you
>>> asked:
>>>
>>> "I use Windows. Came across Linux as another operating system. How
>>> does it compare with Windows and can you use the same applications?"
>>
>> Why only "the same applications"? - this is again limiting the horizon.
>
> Your original question involved "convert each other's codes" -- "the same
> applications" was my analogy to that part of your original questions.

True, but it was just a "throw-in".
>
>>> [Yeah, I took a little bit of liberty with the wording of that question.
>>> It's in the spirit of your original question, though.]
>>>
>>> Answering this question would be MUCH easier if we knew what you were
>>> using Windows to do.
>>
>> This is from your angle. But from my angle I like to find out its full
>> potentials and
>> possibly discover new uses for it and not be limited to a single use. If
>> you think
>> about it is a positive marketing position to take.
>
> Okay, if you want to see the "full potentials", I have a couple of
> references that you may be interested in reviewing. The first is the
> MATLAB and Simulink Based Books page, which includes descriptions of over
> 1200 books that make use of MATLAB and/or Simulink in various ways.
>
> http://www.mathworks.com/support/books/
>
> The second is the product list:
>
> http://www.mathworks.com/products/product_listing/index.html
>
> Each individual product's page contains a description of that product,
> demos and lists of features present in that product, links to the
> documentation for the product, and technical literature and stories about
> how users have used that product in their work.
>
> Of course, if you were, say, working in the field of computational biology
> (just to choose one particular topic at random) then there's a page that
> is targeted toward your particular area of interest.
>
> http://www.mathworks.com/computational-biology/
>
> That way you wouldn't have to weed through all the discussion about topics
> in which you're not interested.

Much obliged. Shall earmark them as my holiday pursuit.

>>> For instance, if you were my mother (who's far from the most
>>> computer-savvy person I know -- sorry Mom) then the answer would be
>>> "Mom, just stick with Windows. You can read web pages on it and play
>>> your games just fine." If you were an IT manager asking for advice
>>> about which OS to run on your servers, then the discussion would be much
>>> more in-depth.
>>>
>>>> To do so would limit the scope of their
>>>> comparative differences to what I was going to use them for.
>>>> I was expecting to get a rounded and objective dissertation.
>>>> The answer from "dpb" did just that. But one respondent
>>>> took a narrow and gladiatorial focus which says more about
>>>> himself than the legitimacy of my question.
>>>
>>> We get people on this newsgroup asking very broad, open-ended questions
>>> relatively frequently. Often, they're people demanding "Here's my
>>> homework, do it for me" and I think that's affected how some of the
>>> gurus respond to those type of broad questions.
>>
>> Why get hot under the collar if the question is provocative? Just don't
>> respond.
>
> As far as I know, everyone posting to this group (except for the spam bots
> trying to sell knock-off shoes or herbal remedies, of course) is human,
> and I can think of very few humans who never get annoyed or frustrated and
> express that annoyance or frustration.

What is missing there is the word, "justification".
But we are digressing from matlab. To be true to my words,
I shall not respond further along this line of discussion.

>
> --
> Steve Lord
> slord@mathworks.com
> comp.soft-sys.matlab (CSSM) FAQ:
> http://matlabwiki.mathworks.com/MATLAB_FAQ
> To contact Technical Support use the Contact Us link on
> http://www.mathworks.com

"Andy " <myfakeemailaddress@gmail.com> wrote in message
news:i5j436$9l0$1@fred.mathworks.com...
>> Yes. Be objective. Confine the discussion to the merits and limitations
>> of the two
>> products. What I am going to use it for does not contribute to the merits
>> of the
>> the product. It only limits the focus. Whereas a broad discussion can
>> reveal
>> properties and capabilities that we may not be aware of and thus can
>> lead
>> to new ideas and usage for the product.
>
> You have to limit your focus. MATLAB and VBA are not inherently good or
> bad. They're both complete programming languages. They have advantages
> and disadvantages in specific areas. There are literally thousands of
> different pros and cons for each language that we could list off, but
> nobody is going to take the time to do that.

Neither was I expecting an exhaustive account nor a narrow focus. But just
what you know about their capabilities and limitations.

>
>
>>Had I just mentioned a particular intended
>>use, I would have missed out on all this information as people would just
>>tailor their response to that
>>particular use.
>
> That's kind of backwards, since I was responding to the only specific uses
> you did mention: developing applications that work with Excel and
> converting code between VBA and MATLAB.

So you ARE aware of what I was going to use it for. Just to say that they
all involve financial modeling - Risk Management, Integrated Financial
System, Return On Investment Analysis, Company Performance Analysis, Stock
Selection Analysis (see attachment somewhere), etc. I don't think they are
of much help.
>
>
>> Why only "the same applications"? - this is again limiting the horizon.
>
> I think this was just an analogy or an example. If somebody were asking
> me to recommend an OS, the most important thing I'd ask them is what
> applications they would want to run. That's a major way to determine the
> correct OS for the user. There are other questions that are, essentially,
> proxies for this (such as my previous comment about gaming).

Here, hardware needs are more important - type of microprocessor, disk and
memory capacity, screen resolution, etc.

> Certainly there are other pros/cons to consider, but you basically
> shouldn't consider most of them. It's not worth your time to learn both
> OSs in sufficient detail to do a full comparison.
> I would argue this is more true for programming languages for multiple
> reasons. As I said before, you are much less tied down to one programming
> language than you are to one OS. In addition, since both VBA and MATLAB
> are complete languages, you can do whatever you want in either, unlike
> with the OS analogy (where some applications really won't run on one OS,
> making that an important distinguishing characteristic).

Nevertheless, just like horses are for courses, each was designed to possess
its own unique quality.
>
>> This is from your angle. But from my angle I like to find out its full
>> potentials and
>> possibly discover new uses for it and not be limited to a single use.
>
> Again, these are complete programming languages. Trying to discover new
> uses is the wrong approach, since each language can do anything you want
> it to. You are more likely to care about things like efficiency of
> workflow, code, and integration with other applications.
>
>
> So, again, it would really help (both you and us) if you would mention
> what your intended uses are. These don't need to be specific. (My
> previous reply was only to the general need for integration with Excel.)

Answered above.

> But even answering the following questions would be useful:
> 1) Will you continue storing your data in Excel?

No, not the raw data. Only the processed data.

> 2) Are you looking to continue integrating heavily with Excel?

Only if the benefits outweigh the efforts.

> Does Excel need to be the front end for your application?

Yes.

> If not, does it need to be part of the application at all?
> 3) What general heading would you put your application under? (Image
> processing? Signal processing? Stock analysis? Etc.)

Which ever heading that is appropriate as there are several of them.

> 4) Do you need to be able to distribute your application?

Currently it is used in-house. But have to encrypt the codes before external
distribution.

>
> There are lots of other important questions, but these will be a good
> start.

Walter Roberson <roberson@hushmail.com> wrote in message <9nQeo.8830$8A2.4277@newsfe22.iad>...
> On 30/08/10 9:38 AM, Rune Allnor wrote:
>
> > That outside the matlab community there is no such thing as
> > a 'vectorized language'.
>
> Rune, if you supply the citations for non-existence, then someone could
> edit the Wikipedia entry on the topic,
> http://en.wikipedia.org/wiki/Array_programming

Or better yet:
http://en.wikipedia.org/wiki/Vectorization_%28computer_science%29

I checked the history, it seems the page predates this thread. :)
And it doesn't even mention MATLAB!

Cheers,
Cris.

> Currently it is used in-house. But have to encrypt the codes before external
> distribution.

Matlab can do that. The "encrypted" code is called p-code (see here http://www.mathworks.com/matlabcentral/newsreader/view_thread/284802).

On 10-09-01 08:09 PM, Tom wrote:

> Neither was I expecting an exhaustive account nor a narrow focus. But
> just what you know about their capabilities and limitations.

Your previous postings were quite specific that you did not want any
assumption made about what might be of interest or not. Very specific details
about some aspects of a language are sometimes crucial to a use, so the only
way to answer your previous unrestricted query would be through exhaustive
account.

> So you ARE aware of what I was going to use it for. Just to say that
> they all involve financial modeling - Risk Management, Integrated
> Financial System, Return On Investment Analysis, Company Performance
> Analysis, Stock Selection Analysis (see attachment somewhere), etc. I
> don't think they are of much help.

Yes, to someone who knows the languages involved, those do help narrow the
aspects that are or are not likely to be of interest to you.

> Nevertheless, just like horses are for courses, each was designed to
> possess its own unique quality.

That would appear to be in accord with my suggestion (to which you did not
reply) that it would be easier if you were to ask about the traditional
strength and weaknesses of the languages.

>> Does Excel need to be the front end for your application?
>
> Yes.

I am not entirely clear at the moment as to whether you would intend to do the
graphics in Excel (VBA) or in Matlab or in "whatever is easiest at the time" ?

 > Neither was I expecting an exhaustive account nor a narrow focus. But
 > just what you know about their capabilities and limitations.

At least three of the regulars have written about Matlab more than ten
thousand times. "Just what you know" gets to be a pretty big list after awhile!



A brief comparison of Matlab and C 89 (C in general is up to C99 + Technical
Report 6 or so, but I know C 89 better):

C has these features or concepts that Matlab does not:

- medium-strong static variable typing (with automatic conversions and with
type escaping mechanisms, but no enforcement that types are used consistently
between different compilation units); types must be pre-declared for all
variables and must be pre-declared for all parameters of a routine; the type
of a variable cannot change at run time

- named numeric constants; arbitrary structures can be marked constant after
initialization; parameters to functions and imported variables can be marked
as not to be changed during the lexical scope

- block-scoped variables and constants

- variable number of arguments to routines requires special handling and is
easy to get wrong

- a repetition control structure whose test is at the end

- a repetition control structure with an initialization expression, a
termination test expression, and a pre-repetition expression

- goto, and statement labels (within the same function)

- "pointers"

- ability to read and write blocks of memory at byte level

- unions

- structures can be defined down to bit widths (but exact alignment cannot be
specified)

- static type creation

- octal and hexadecimal numeric constants; octal, decimal and hexadecimal
character initialization (including within strings)

- machine-native character strings

- assignment operators within expressions

- very limited error detection and trapping

- lexical inclusion of source

- conditional parsing and compilation of source

- lexical constants

- a preprocessing language that can greatly change the potential appearance of
source

- "name-spaces" (not in C89, but available in C99)

- ability to generate small stand-alone executables or executables that only
call upon standard OS routines


Matlab has these features or concepts that C 89 does not:

- dynamic typing of all variables; limited automatic type conversion; no way
to declare the expected or required types for variables or routine parameters

- run-time type detection (types of variables can be queried at run time)

- variable number of arguments to routines is allowed, pervasive, and simple

- routines may return multiple arguements

- ability to dynamically construct and evaluate code

- extended (UTF-16) characters are standard internally

- vector and array constructors; vector and array operators

- (limited) ability to read from or write to variables at different scoping levels

- a mechanism for an array to contain mixed sizes and types of objects

- built in graphics

- a built-in debugger

- a provided integrated development environment

- generalized error detection and trapping

- built in support for complex numbers (C89 does not offer this, but C99 does.)

- shared variables and closures

- anonymous functions

- I/O operators can operate at the bit level

- Some integration with Java, .NET, and ActiveX

- built-in object-oriented programming

- a much larger library of provided mathematical and engineering tasks

- optional (additional-cost) parallel processing

- optional (additional-cost) libraries of common tasks in a number of
different domains


If you examine these lists, you will see that the two languages really do not
differ substantially in fundamental facilities. At the implementation level,
the single biggest difference is not the array/vector facilities in Matlab,
nor (really) the static vs dynamic typing: the biggest single difference is
the much less obvious generalized error detection and trapping in Matlab,
which C cannot (reliably) emulate.


Matlab's support for graphics and for the integrated development environment
come at the expense of speed and portability and size. I used to do a lot of
non-graphical C development work on a 256 kilobyte 10 megahertz CISC machine
(roughly 2 MHz RISC equivalent), but Matlab's graphics and IDE were unusable
on a 8 gigabyte 250 MHz RISC machine.


Beyond that, the great majority of things you can do in C, you can also do in
Matlab, and the great majority of things you can do in Matlab could be
translated to C without conceptual difficulty (but possibly with a lot of
recoding of numeric libraries.)


Possibly the greatest deficiency of Matlab compared to C (well, other than
memory size and speed) that Matlab does not provide any convenient built-in
mechanisms for direct access to libraries written to standard interfaces. If
one examines the strict definition of C, then in theory C does not really have
a strong interface either -- but in practice nearly everyone standardized on
interfaces that could be used from C without much trouble, whereas the Matlab
external interface is inherently Mathworks-specific.

"Walter Roberson" <roberson@hushmail.com> wrote in message
news:i5p9en$b8s$1@canopus.cc.umanitoba.ca...
> On 10-09-01 08:09 PM, Tom wrote:
>
>> Neither was I expecting an exhaustive account nor a narrow focus. But
>> just what you know about their capabilities and limitations.
>
> Your previous postings were quite specific that you did not want any
> assumption made about what might be of interest or not. Very specific
> details about some aspects of a language are sometimes crucial to a use,
> so the only way to answer your previous unrestricted query would be
> through exhaustive account.

No obligation was intended nor a zealous response expected.
>
>> So you ARE aware of what I was going to use it for. Just to say that
>> they all involve financial modeling - Risk Management, Integrated
>> Financial System, Return On Investment Analysis, Company Performance
>> Analysis, Stock Selection Analysis (see attachment somewhere), etc. I
>> don't think they are of much help.
>
> Yes, to someone who knows the languages involved, those do help narrow the
> aspects that are or are not likely to be of interest to you.
>
>> Nevertheless, just like horses are for courses, each was designed to
>> possess its own unique quality.
>
> That would appear to be in accord with my suggestion (to which you did not
> reply) that it would be easier if you were to ask about the traditional
> strength and weaknesses of the languages.
>
>>> Does Excel need to be the front end for your application?
>>
>> Yes.
>
> I am not entirely clear at the moment as to whether you would intend to do
> the graphics in Excel (VBA) or in Matlab or in "whatever is easiest at the
> time" ?

Graphics capability in Excel is quite primitive compared to other dedicated
software for
Stock Analysis. I chose Excel mainly for its mathematical capabilities which
I believe
Matlab excels in. I haven't heard any comment about the graphics of Matlab.
For me
to switch over to Matlab, the benefits must outweigh the efforts or else
future projects
demand functionalities that are absent in Excel.
>
> > Neither was I expecting an exhaustive account nor a narrow focus. But
> > just what you know about their capabilities and limitations.
>
> At least three of the regulars have written about Matlab more than ten
> thousand times. "Just what you know" gets to be a pretty big list after
> awhile!

Only wish they had documented them in their web sites like some of the Excel
gurus do.

> A brief comparison of Matlab and C 89 (C in general is up to C99 +
> Technical Report 6 or so, but I know C 89 better):
>
> C has these features or concepts that Matlab does not:
>
> - medium-strong static variable typing (with automatic conversions and
> with type escaping mechanisms, but no enforcement that types are used
> consistently between different compilation units); types must be
> pre-declared for all variables and must be pre-declared for all parameters
> of a routine; the type of a variable cannot change at run time
>
> - named numeric constants; arbitrary structures can be marked constant
> after initialization; parameters to functions and imported variables can
> be marked as not to be changed during the lexical scope
>
> - block-scoped variables and constants
>
> - variable number of arguments to routines requires special handling and
> is easy to get wrong
>
> - a repetition control structure whose test is at the end
>
> - a repetition control structure with an initialization expression, a
> termination test expression, and a pre-repetition expression
>
> - goto, and statement labels (within the same function)
>
> - "pointers"
>
> - ability to read and write blocks of memory at byte level
>
> - unions
>
> - structures can be defined down to bit widths (but exact alignment cannot
> be specified)
>
> - static type creation
>
> - octal and hexadecimal numeric constants; octal, decimal and hexadecimal
> character initialization (including within strings)
>
> - machine-native character strings
>
> - assignment operators within expressions
>
> - very limited error detection and trapping
>
> - lexical inclusion of source
>
> - conditional parsing and compilation of source
>
> - lexical constants
>
> - a preprocessing language that can greatly change the potential
> appearance of source
>
> - "name-spaces" (not in C89, but available in C99)
>
> - ability to generate small stand-alone executables or executables that
> only call upon standard OS routines
>
>
> Matlab has these features or concepts that C 89 does not:
>
> - dynamic typing of all variables; limited automatic type conversion; no
> way to declare the expected or required types for variables or routine
> parameters
>
> - run-time type detection (types of variables can be queried at run time)
>
> - variable number of arguments to routines is allowed, pervasive, and
> simple
>
> - routines may return multiple arguements
>
> - ability to dynamically construct and evaluate code
>
> - extended (UTF-16) characters are standard internally
>
> - vector and array constructors; vector and array operators
>
> - (limited) ability to read from or write to variables at different
> scoping levels
>
> - a mechanism for an array to contain mixed sizes and types of objects
>
> - built in graphics
>
> - a built-in debugger
>
> - a provided integrated development environment
>
> - generalized error detection and trapping
>
> - built in support for complex numbers (C89 does not offer this, but C99
> does.)
>
> - shared variables and closures
>
> - anonymous functions
>
> - I/O operators can operate at the bit level
>
> - Some integration with Java, .NET, and ActiveX
>
> - built-in object-oriented programming
>
> - a much larger library of provided mathematical and engineering tasks
>
> - optional (additional-cost) parallel processing
>
> - optional (additional-cost) libraries of common tasks in a number of
> different domains
>
>
> If you examine these lists, you will see that the two languages really do
> not differ substantially in fundamental facilities. At the implementation
> level, the single biggest difference is not the array/vector facilities in
> Matlab, nor (really) the static vs dynamic typing: the biggest single
> difference is the much less obvious generalized error detection and
> trapping in Matlab, which C cannot (reliably) emulate.
>
>
> Matlab's support for graphics and for the integrated development
> environment come at the expense of speed and portability and size. I used
> to do a lot of non-graphical C development work on a 256 kilobyte 10
> megahertz CISC machine (roughly 2 MHz RISC equivalent), but Matlab's
> graphics and IDE were unusable on a 8 gigabyte 250 MHz RISC machine.
>
>
> Beyond that, the great majority of things you can do in C, you can also do
> in Matlab, and the great majority of things you can do in Matlab could be
> translated to C without conceptual difficulty (but possibly with a lot of
> recoding of numeric libraries.)
>
>
> Possibly the greatest deficiency of Matlab compared to C (well, other than
> memory size and speed) that Matlab does not provide any convenient
> built-in mechanisms for direct access to libraries written to standard
> interfaces. If one examines the strict definition of C, then in theory C
> does not really have a strong interface either -- but in practice nearly
> everyone standardized on interfaces that could be used from C without much
> trouble, whereas the Matlab external interface is inherently
> Mathworks-specific.

Interesting reading. You could point out these limitations to the developer
of Matlab and request that they be implemented in their future upgrades.

"Gary" <grubin698@gmail.com> wrote in message
news:i5o6lk$8pm$1@fred.mathworks.com...
>
>> Currently it is used in-house. But have to encrypt the codes before
>> external distribution.
>
> Matlab can do that. The "encrypted" code is called p-code (see here
> http://www.mathworks.com/matlabcentral/newsreader/view_thread/284802).

It is interesting to know. Wonder how strong is the encryption and whether
anyone has broken it?

Caveat: I'm only a casual user of Excel, so I don't know if VBA runs into the same limitations as "standard" Excel.

In my experience, one of the big benefits of Matlab over Excel is that Matlab is not limited by the maximum number of rows that can be loaded into memory. We often deal with quite long CSV files, and are regularly told be Excel that our data sets are too long to load (or plot).

Gary

"Tom" <tclim36@gmail.com> wrote in message
news:tG6go.5414$FH2.3986@viwinnwfe02.internal.bigpond.com...
> "Walter Roberson" <roberson@hushmail.com> wrote in message
> news:i5p9en$b8s$1@canopus.cc.umanitoba.ca...
>> On 10-09-01 08:09 PM, Tom wrote:

*snip*

>> At least three of the regulars have written about Matlab more than ten
>> thousand times. "Just what you know" gets to be a pretty big list after
>> awhile!
>
> Only wish they had documented them in their web sites like some of the
> Excel
> gurus do.

Well, the product documentation, which contains a lot of the information
I've posted to the newsgroup, is located on our support website:

http://www.mathworks.com/support/

and as of sometime last night, documentation for releases prior to the
current release (back to release R13SP2, which was released in 2004) is also
available.

http://www.mathworks.com/help/doc-archives.html

There's also the Newsgroup interface and the File Exchange code repository
on MATLAB Central:

http://www.mathworks.com/matlabcentral/

--
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ
To contact Technical Support use the Contact Us link on
http://www.mathworks.com

"Tom" <tclim36@gmail.com> wrote in message <w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com>...
> I use VBA to develop applications for Excel. Came across Matlab
> as another high level language. How does it compare with VBA
> and can you convert each other's codes? Thanks for your advice.
>
> Tom

VBA is good but the thing is that matlab has much more efficient ways to solve a problem...

In VBA you need to do the calculations in MATLAB some fucntions are ready or easy to write;

now you tell me which one is better...

"Tom" <tclim36@gmail.com> wrote in message <tG6go.5414$FH2.3986@viwinnwfe02.internal.bigpond.com>...
<snip_
>
> Interesting reading. You could point out these limitations to the developer
> of Matlab and request that they be implemented in their future upgrades.

From your first post stating that you "Came across Matlab as another high-level language" to this?

I think this fits the standard definition of trolling quite well.

On 03/09/10 8:42 AM, Tom wrote:
> "Gary" <grubin698@gmail.com> wrote in message
> news:i5o6lk$8pm$1@fred.mathworks.com...
>>
>>> Currently it is used in-house. But have to encrypt the codes before
>>> external distribution.
>>
>> Matlab can do that. The "encrypted" code is called p-code (see here
>> http://www.mathworks.com/matlabcentral/newsreader/view_thread/284802).
>
> It is interesting to know. Wonder how strong is the encryption and
> whether anyone has broken it?

pcode files are not encrypted; they are merely the binary files that are
the same as you would get internally from the Just In Time compiler.

If you use the "Matlab Compiler" (which isn't really a compiler), then
there is an option to apply AES encryption to the Component Technology
File repository.

"Walter Roberson" <roberson@hushmail.com> wrote in message
news:RU8go.881$EV4.2@newsfe16.iad...
> On 03/09/10 8:42 AM, Tom wrote:
>> "Gary" <grubin698@gmail.com> wrote in message
>> news:i5o6lk$8pm$1@fred.mathworks.com...
>>>
>>>> Currently it is used in-house. But have to encrypt the codes before
>>>> external distribution.
>>>
>>> Matlab can do that. The "encrypted" code is called p-code (see here
>>> http://www.mathworks.com/matlabcentral/newsreader/view_thread/284802).
>>
>> It is interesting to know. Wonder how strong is the encryption and
>> whether anyone has broken it?
>
> pcode files are not encrypted; they are merely the binary files that are
> the same as you would get internally from the Just In Time compiler.
>
> If you use the "Matlab Compiler" (which isn't really a compiler), then
> there is an option to apply AES encryption to the Component Technology
> File repository.

Does it mean that only you can retrieve the original codes and no one else?
I do have a third party product for Excel that has a strong encryption from
which the codes cannot be recovered. So I keep a copy of my original codes
before applying the encryption.

"Gary" <grubin698@gmail.com> wrote in message
news:i5qui7$a2e$1@fred.mathworks.com...
> Caveat: I'm only a casual user of Excel, so I don't know if VBA runs into
> the same limitations as "standard" Excel.
>
> In my experience, one of the big benefits of Matlab over Excel is that
> Matlab is not limited by the maximum number of rows that can be loaded
> into memory. We often deal with quite long CSV files, and are regularly
> told be Excel that our data sets are too long to load (or plot).

I started with Excel11 (R=65,536; C=256). To overcome column limitation I
resorted to segmentation. Current Excel12 (R=1,048576; C=16,384) is
adequate but I can't use my old menu-bar. So I continue to use Excel11.
>
> Gary

You have been very helpful indeed, Steve. Your contributions are much
appreciated.

"Steven_Lord" <slord@mathworks.com> wrote in message
news:i5qvjb$h9j$1@fred.mathworks.com...
>
>
> "Tom" <tclim36@gmail.com> wrote in message
> news:tG6go.5414$FH2.3986@viwinnwfe02.internal.bigpond.com...
>> "Walter Roberson" <roberson@hushmail.com> wrote in message
>> news:i5p9en$b8s$1@canopus.cc.umanitoba.ca...
>>> On 10-09-01 08:09 PM, Tom wrote:
>
> *snip*
>
>>> At least three of the regulars have written about Matlab more than ten
>>> thousand times. "Just what you know" gets to be a pretty big list after
>>> awhile!
>>
>> Only wish they had documented them in their web sites like some of the
>> Excel
>> gurus do.
>
> Well, the product documentation, which contains a lot of the information
> I've posted to the newsgroup, is located on our support website:
>
> http://www.mathworks.com/support/
>
> and as of sometime last night, documentation for releases prior to the
> current release (back to release R13SP2, which was released in 2004) is
> also available.
>
> http://www.mathworks.com/help/doc-archives.html
>
> There's also the Newsgroup interface and the File Exchange code repository
> on MATLAB Central:
>
> http://www.mathworks.com/matlabcentral/
>
> --
> Steve Lord
> slord@mathworks.com
> comp.soft-sys.matlab (CSSM) FAQ:
> http://matlabwiki.mathworks.com/MATLAB_FAQ
> To contact Technical Support use the Contact Us link on
> http://www.mathworks.com

"Mark Shore" <mshore@magmageosciences.ca> wrote in message
news:i5r30q$1pp$1@fred.mathworks.com...
> "Tom" <tclim36@gmail.com> wrote in message
> <tG6go.5414$FH2.3986@viwinnwfe02.internal.bigpond.com>...
> <snip_
>>
>> Interesting reading. You could point out these limitations to the
>> developer
>> of Matlab and request that they be implemented in their future upgrades.
>
> From your first post stating that you "Came across Matlab as another
> high-level language" to this?
>
> I think this fits the standard definition of trolling quite well.

Standard? You mean a narrow minded standard coming from a reclusive
arm-chair critic, bereft of positive contribution except sarcasm.

"Tom" <tclim36@gmail.com> wrote in message <x8igo.5312$Yv.5052@viwinnwfe01.internal.bigpond.com>...
> "Mark Shore" <mshore@magmageosciences.ca> wrote in message
> news:i5r30q$1pp$1@fred.mathworks.com...
> > "Tom" <tclim36@gmail.com> wrote in message
> > <tG6go.5414$FH2.3986@viwinnwfe02.internal.bigpond.com>...
> > <snip_
> >>
> >> Interesting reading. You could point out these limitations to the
> >> developer
> >> of Matlab and request that they be implemented in their future upgrades.
> >
> > From your first post stating that you "Came across Matlab as another
> > high-level language" to this?
> >
> > I think this fits the standard definition of trolling quite well.
>
> Standard? You mean a narrow minded standard coming from a reclusive
> arm-chair critic, bereft of positive contribution except sarcasm.


From Wikipedia: "In The Art of Trolling, published on the web, it is suggested that in Usenet usage, a "troll" is not a grumpy monster that lives beneath a bridge accosting passers by, but rather a provocative posting to a news group intended to produce a large volume of frivolous responses. The content of a "troll" posting generally falls into several areas. It may consist of an apparently foolish contradiction of common knowledge, a deliberately offensive insult to the readers of the news group or a broad request for trivial follow-up postings."

I implied the last meaning. And if I were naive enough to ask a question on a C++ forum by stating I'd just heard about the language and what are the differences between OOP C++ and MATLAB, and then follow it up within a day by suggesting fundamental changes to C++, I would not expect enthusiastic replies either.

Don't take it to mean that I am here to criticise Matlab but rather to learn
from users in what way they find
the language useful in comparison to VBA so that I may be a convert. The
number of functions are quite
sufficient with help on how to use them. Also it allows users to write their
own. Some of the limitations
I find are:
1. When testing your codes you can't use the pause command (unlike the
earlier version) and pick up from there.
2. There is a limit to the total length of your codes.
3. A lengthy way to make use of the reference to a particular cell.
4. Graphics are just basic.
5. Harder to create your own menu with the new version Excel12.

You said,
> VBA is good but the thing is that matlab has much more efficient ways to
> solve a problem...
1. Can you give a simple example?
2. Also, can Matlab automatically generate the codes while you carry out a
simple task like copy and paste?


"mer KAYA" <e170531@metu.edu.tr> wrote in message
news:i5r24h$3ck$1@fred.mathworks.com...
> "Tom" <tclim36@gmail.com> wrote in message
> <w18eo.4732$Yv.4180@viwinnwfe01.internal.bigpond.com>...
>> I use VBA to develop applications for Excel. Came across Matlab
>> as another high level language. How does it compare with VBA
>> and can you convert each other's codes? Thanks for your advice.
>>
>> Tom
>
> VBA is good but the thing is that matlab has much more efficient ways to
> solve a problem...
> In VBA you need to do the calculations in MATLAB some fucntions are ready
> or easy to write;
> now you tell me which one is better...

"Mark Shore" <mshore@magmageosciences.ca> wrote in message
news:i5sf0o$72l$1@fred.mathworks.com...
> "Tom" <tclim36@gmail.com> wrote in message
> <x8igo.5312$Yv.5052@viwinnwfe01.internal.bigpond.com>...
>> "Mark Shore" <mshore@magmageosciences.ca> wrote in message
>> news:i5r30q$1pp$1@fred.mathworks.com...
>> > "Tom" <tclim36@gmail.com> wrote in message
>> > <tG6go.5414$FH2.3986@viwinnwfe02.internal.bigpond.com>...
>> > <snip_
>> >>
>> >> Interesting reading. You could point out these limitations to the
>> >> developer
>> >> of Matlab and request that they be implemented in their future
>> >> upgrades.
>> >
>> > From your first post stating that you "Came across Matlab as another
>> > high-level language" to this?
>> >
>> > I think this fits the standard definition of trolling quite well.
>>
>> Standard? You mean a narrow minded standard coming from a reclusive
>> arm-chair critic, bereft of positive contribution except sarcasm.
>
>
> From Wikipedia: "In The Art of Trolling, published on the web, it is
> suggested that in Usenet usage, a "troll" is not a grumpy monster that
> lives beneath a bridge accosting passers by, but rather a provocative
> posting to a news group intended to produce a large volume of frivolous
> responses. The content of a "troll" posting generally falls into several
> areas. It may consist of an apparently foolish contradiction of common
> knowledge, a deliberately offensive insult to the readers of the news
> group or a broad request for trivial follow-up postings."
>
> I implied the last meaning. And if I were naive enough to ask a question
> on a C++ forum by stating I'd just heard about the language and what are
> the differences between OOP C++ and MATLAB, and then follow it up within a
> day by suggesting fundamental changes to C++, I would not expect
> enthusiastic replies either.

I hope you can comprehend the fine nuances below. You don't need to have a
legal mind.
1. Making me out to be the agent seeking a "fundamental change", when my
contribution is no more than making a suggestion to a third party to have
his idea known.
2. An upgrade is not a fundamental change.

But how clever is it to come up with a "standard definition" that fits ones
narrow mind? What follow answer that question.
Here is another one:
What Is A Troll? The WWW gives this as a definition:
troll v.,n. To utter a posting on Usenet designed to attract predictable
responses or flames. Derives from the phrase "trolling for newbies"; which
in turn comes from mainstream "trolling";, a style of fishing in which one
trails bait through a likely spot hoping for a bite.

One other:
What is a snake in the grass?
An elongated reptile that hides in toll grasses. Known to sink its fangs
into innocent passers-by often unprovoked. Also a term used to refer to a
person with similar characteristics who contributes nothing positive except
to show his or her smugness.
 

On 03/09/10 8:08 PM, Tom wrote:
> "Walter Roberson" <roberson@hushmail.com> wrote in message

>> If you use the "Matlab Compiler" (which isn't really a compiler), then
>> there is an option to apply AES encryption to the Component Technology
>> File repository.

> Does it mean that only you can retrieve the original codes and no one else?
> I do have a third party product for Excel that has a strong encryption from
> which the codes cannot be recovered. So I keep a copy of my original codes
> before applying the encryption.

The Matlab Compiler does not copy the source in to the generated
executable; it reduces the source into pcode (threaded interpreted code)
and the pcode is what is put in the object. The AES encryption (with
unknown random key, I believe) protects even further as makes reverse
engineering even more difficult (as otherwise it could be attacked
through differential cryptography techniques.)

Anyhow, when you use the Matlab Compiler, make sure you retain a clean
readable copy of the code.

"Walter Roberson" <roberson@hushmail.com> wrote in message
news:b1dho.172640$1F6.49076@newsfe01.iad...
> On 03/09/10 8:08 PM, Tom wrote:
>> "Walter Roberson" <roberson@hushmail.com> wrote in message
>
>>> If you use the "Matlab Compiler" (which isn't really a compiler), then
>>> there is an option to apply AES encryption to the Component Technology
>>> File repository.
>
>> Does it mean that only you can retrieve the original codes and no one
>> else?
>> I do have a third party product for Excel that has a strong encryption
>> from
>> which the codes cannot be recovered. So I keep a copy of my original
>> codes
>> before applying the encryption.
>
> The Matlab Compiler does not copy the source in to the generated
> executable; it reduces the source into pcode (threaded interpreted code)
> and the pcode is what is put in the object. The AES encryption (with
> unknown random key, I believe) protects even further as makes reverse
> engineering even more difficult (as otherwise it could be attacked through
> differential cryptography techniques.)
>
> Anyhow, when you use the Matlab Compiler, make sure you retain a clean
> readable copy of the code.

Shall do when I get to that stage. Thanks.