Wolfram - Page 2

ELSID · ‎Aug 03, 2009

Mona,
I received an e-mail that I find disturbing. Apparently our e-mails are being "bot" d out of the collab

"Hi. I happened to stumble on your post over at the Mathcad user forum and wondered if you've had a chance to check out Wolfram|Alpha.

...[Deleted]

Thanks,

Mary Richmond
Wolfram Research"

No knowledge can be certain if it is not based upon mathematics or upon some other
knowledge which is itself based upon the mathematical sciences. -- Leonardo da Vinci

ptc-1368288 · ‎Oct 21, 2009

On 10/21/2009 9:54:10 PM, rm_ulrich wrote:
>On 10/21/2009 8:30:03 PM, jmG wrote:
>>Here are 8 random data set from "one of
>>the clowns".
>
>One clown to another, can you tell me
>how you inserted the red separator lines
>into the document? I've looked in vain
>for ways insert a line or box using
>Mathcad 14.
>
>Thanks,
>___________________________________

There is no red separator in my work sheet.
A brown metafile Excel line, simply.

jmG

rm_ulrich · ‎Oct 21, 2009

On 10/21/2009 10:21:22 PM, jmG wrote:
>A brown metafile Excel line, simply.

Ok, fine, it's brown. But how do you create/insert it?

Thanks,
R

ptc-1368288 · ‎Oct 21, 2009

On 10/21/2009 10:34:56 PM, rm_ulrich wrote:
>On 10/21/2009 10:21:22 PM, jmG wrote:
>>A brown metafile Excel line, simply.
>
>Ok, fine, it's brown. But how do you
>create/insert it?
>
>Thanks,
>R
______________________________

Draw the line in Excel, copy, paste in Mathcad.

jmG

TomGutman · ‎Oct 22, 2009

WHen I want a simple separator line, I just lean on the undescore key for a bit.
__________________
� � � � Tom Gutman

rm_ulrich · ‎Oct 22, 2009

On 10/22/2009 2:05:38 AM, Tom_Gutman wrote:
>WHen I want a simple separator
>line, I just lean on the
>undescore key for a bit.
>__________________
>� � � � Tom Gutman

But what I really wanted to do was draw a box or a simple block diagram. Now I can do it, if a bit kludgy. A quick experiment shows that fields underneath an unfilled box are visible, so I can have outputs shown there, so I'm 90% to where I wanted to be, which is good: Intergraph wants a ton of money for a single seat of SmartSketch.

Meanwhile, I have a call back waiting for me from Mathcad tech support, to track down rumors that there are more elegant solutions involving autocad or autodesk or whatever it is called.

R

ptc-1368288 · ‎Oct 22, 2009

What is "SmartSketch" and Autocad doing here ?
This thread is about the "two clowns" fitting challenge ... MMMM
MMM = Mathcad/Maple/Mathematica/Matlab.
No idea what your text box is for.

jmG

mikej1980 · ‎Oct 22, 2009

On 10/21/2009 8:30:03 PM, jmG wrote:
>On 10/21/2009 2:59:04 PM, beetle wrote:
>>Can you clowns come up with a
>>ONE single curve fitting
>>problem
>>that can't be done better in
>>Maple, Mathematica or Matlab?
>>
>>I have access to all 4 Ms. My
>>bet is that you have no idea
>>of how
>>far behind Mathcad is ...
>>Post your toughest curve
>>fitting problem and I'll post
>>solutions
>>in the other 3 Ms.
>______________________________
>
>Here are 8 random data set from "one of
>the clowns". You return the model
>function fitted. Curve fitting in this
>collab has always been done by model
>functions. Don't care posting
>interpolated splined or else data,
>Mathcad can do, but the challenge to
>data fitter is more than just piecewise
>interpolation and compact in data table
>... the specification to the other 3M's
>is the model function for each, he best
>fit model.
>
>jmG
>
>

yeah too cute by half. So you want me to manually write down the data in each of your data sets do you.

Provide the data you want me to fit and the model function. If you are genuine you will do this. The other 3Ms don't read mathcad binary and even after converting to mathcad xml I got some xml errors trying to import into other Ms.

provide data sets that are readable in other programs (simple text or cvs will do fine or xls) and a model function and I will show you how much better the other Ms are at curve fitting.

To make it a better challenge give me a model with constraints.

ptc-1368288 · ‎Oct 22, 2009

My dear collab:

Only the last one yellow you will do in Matlab. In this forum since it started "Collab started 30 Aug 1999", several 100's of data fitting were done. I have seen all and done most of hem. This requires a library of models [guess how huge !], some have taken if not weeksssss, week. Fitting is more than few ping-pong pieces of maths, it is an art [F.B. Hildebrand]. Mathcad is the best tool box for fitting/modeling. It's no game plugging the model in Mathematica and get the fitting coefficients, the game is the model itself. In this collab, fitting expertise is immense, still missing some automatic search for the initial guesses and few more bits. This one attached shows how involved a model can be. The model is not explained or otherwise derived, just plug at best in your 3M's. The fit was done in several collaboration. With the 3 other M's you will have the problem of exchanging the work sheets. Years ago, the 3M's had no work sheet exchange in their user forum, has it changed ?

jmG

PS: "Save as" version 11 for a much larger community to read.
I'm at version 11.2a

mikej1980 · ‎Oct 22, 2009

On 10/22/2009 2:39:59 PM, jmG wrote:
>My dear collab:
>
>Only the last one yellow you
>will do in Matlab. In this
>forum since it started "Collab
>started 30 Aug 1999", several
>100's of data fitting were
>done. I have seen all and done
>most of hem. This requires a
>library of models [guess how
>huge !], some have taken if
>not weeksssss, week. Fitting
>is more than few ping-pong
>pieces of maths, it is an art
>[F.B. Hildebrand]. Mathcad is
>the best tool box for
>fitting/modeling. It's no game
>plugging the model in
>Mathematica and get the
>fitting coefficients, the game
>is the model itself. In this
>collab, fitting expertise is
>immense, still missing some
>automatic search for the
>initial guesses and few more
>bits. This one attached shows
>how involved a model can be.
>The model is not explained or
>otherwise derived, just plug
>at best in your 3M's. The fit
>was done in several
>collaboration. With the 3
>other M's you will have the
>problem of exchanging the work
>sheets. Years ago, the 3M's
>had no work sheet exchange in
>their user forum, has it
>changed ?
>
>jmG
>
>PS: "Save as" version 11 for a
>much larger community to read.
>I'm at version 11.2a

I'm making allowances for the fact that english doesn't seem to be your first language so let me say it one more time:

Can you please supply me the data. Do you expect me to hand copy 600 pairs of data from you file!! If you have that data then why not provide it. The "challenge" is how well other Ms can do the curve fitting not how long you can test my patience.

I don't need to exchange worksheets between any of the Ms. All I need is the raw data. All the Ms will import it.

Data please ...

ptc-1368288 · ‎Oct 22, 2009

... let me say it one more time:

Can you please supply me the data.

==> The work sheet is c/w the raw data
==> Otherwise, how do you think you would read the fit below ?
==> Click on the *.gif attached

Do you expect me to hand copy 600 pairs of data from you file!!

==> I expect nothing, you defied "two clowns"
==> ... pass the data to the 3M's the way you want, or the way they can take. The Mathcad data table, spool in file and gear the 3M's to read.

If you have that data then why not provide it.

==> From your repost, the data are there as you confirmed 600 !

The "challenge" is how well other Ms can do the curve fitting not how long you can test my patience.

==> The first challenge is for you to pass the data from a Mathcad work sheet to the 3M's. Read the Mathcad qs from examples how to save data table in file, in file that the 3M's will read.

I don't need to exchange worksheets between any of the Ms. All I need is the raw data. All the Ms will import it.

Data please ...

"All I need is the raw data"

==>You have as you confirmed 600 pairs of X's: Y's

I can't tell if your Mathcad version (supposed above 11) does not read a data table from version 11. Do you have a genuine Mathcad from purchase or some kind of crack ? On that kind of dilemma, hope some collab will confirm if they read the data from my work sheet. If you figure "data table" is not english enough for either of the 3M's, go with their rules.

jmG

mikej1980 · ‎Oct 22, 2009

On 10/22/2009 2:39:59 PM, jmG wrote:
>My dear collab:
>
>Only the last one yellow you
>will do in Matlab. In this
>forum since it started "Collab
>started 30 Aug 1999", several
>100's of data fitting were
>done. I have seen all and done
>most of hem. This requires a
>library of models [guess how
>huge !], some have taken if
>not weeksssss, week. Fitting
>is more than few ping-pong
>pieces of maths, it is an art
>[F.B. Hildebrand]. Mathcad is
>the best tool box for
>fitting/modeling. It's no game
>plugging the model in
>Mathematica and get the
>fitting coefficients, the game
>is the model itself. In this
>collab, fitting expertise is
>immense, still missing some
>automatic search for the
>initial guesses and few more
>bits. This one attached shows
>how involved a model can be.
>The model is not explained or
>otherwise derived, just plug
>at best in your 3M's. The fit
>was done in several
>collaboration. With the 3
>other M's you will have the
>problem of exchanging the work
>sheets. Years ago, the 3M's
>had no work sheet exchange in
>their user forum, has it
>changed ?
>
>jmG
>
>PS: "Save as" version 11 for a
>much larger community to read.
>I'm at version 11.2a

I should have mentioned that of course the model is the most
important but that is not something for any of the Ms that is for
the scientist of engineer with specialized knowledge of the
experiment he is getting his data from!

You prompted my post by saying Mathematica is not geared for
curve fitting and then went off about how great Mathcad is. My
response is that apart from units Mathcad is miles (no pun
intended) behind the other Ms.

The way to test your assertion about Mathcad vs other Ms for
curve fitting is to supply a data set and a model. If any sane
contributors here can think of other or better ways I am open to
suggestions.

So far all you have done is send data buried in (horribly
rendered) mathcad binary worksheets -- i.e. data that cannot be
uploaded and tested by other Ms.

I am saying Mathcad is way behind. If anyone wants to challenge
that then provide some data and a model to test.

StuartBruff · ‎Oct 22, 2009

On 10/22/2009 4:17:21 PM, beetle wrote:
== I am saying Mathcad is way behind. If anyone wants to challenge that then provide some data and a model to test.

Which is something I have been saying/predicting, Cassandra-like (but not Cassandra-looking), for over 5 years. Sadly, I'm not too sure the concepts of resting-on-laurels or tortoise & hare, or better yet, the Boyd Cycle form part of the Mathcad mindset. Mathcad ought to have been so far ahead of where it is in terms of interface and capability.

Stuart

IRstuff · ‎Oct 22, 2009

On 10/22/2009 4:17:30 PM, beetle wrote:
>So far all you have done is send data
>buried in (horribly
>rendered) mathcad binary worksheets --
>i.e. data that cannot be
>uploaded and tested by other Ms.

While I would rarely defend jmg and his antics, it seems a bit disingenuous to claim that his data is inaccessible. Select any column, copy selection, and paste where you want.

TTFN,
Eden

TheodoreM.Bones · ‎Oct 22, 2009

Can any other aid produce the model and curve fit of data to the deProny algorith? It is on MathCad per my submission long ago. I have scanned lots of PDFs on the 'deProny Equation' and found not one in the Hankel upward sloping diagonal form for either equally or geometrically spaced x-data. All the other aids seem to avoid the entire subject.

mikej1980 · ‎Oct 23, 2009

On 10/22/2009 5:09:30 PM, eden_mei wrote:
>On 10/22/2009 4:17:30 PM, beetle wrote:
>>So far all you have done is send data
>>buried in (horribly
>>rendered) mathcad binary worksheets --
>>i.e. data that cannot be
>>uploaded and tested by other Ms.
>
>
>While I would rarely defend jmg and his
>antics, it seems a bit disingenuous to
>claim that his data is inaccessible.
>Select any column, copy selection, and
>paste where you want.
>
>TTFN,
>Eden

I tried what I could to extract the data. I was unable to copy and
paste columns, only individual entires and like I said I am not
going to do that 2 X 600 times. This guy has the data in a
format that can be provided but chooses to be obstructive. If he
believes that the other Ms can't do as well as mathcad he'd
actually be more than happy to provide the data I would have
thought, instead he's being disingenuous. ...but further reading
it looks like he is the forum kook.

The data and model looked like a step decay process (like a
laser flash experiment in chemistry) which is trivial to model.

IRstuff · ‎Oct 23, 2009

Talk is cheap; I copied out the 3 biggest sets of data. Took all of 20 seconds.

TTFN,
Eden

RichardJ · ‎Oct 23, 2009

On 10/23/2009 11:02:49 AM, beetle wrote:

>I tried what I could to extract the
>data. I was unable to copy and
>paste columns, only individual entires
>and like I said I am not
>going to do that 2 X 600 times.

Select all the entries you want, right click, "Copy Selection".

One thing to be aware of is that the data is only copied to the displayed precision (I consider this a bug). So I always go to "properties" first and change the precision to 15.

Richard

StuartBruff · ‎Oct 23, 2009

On 10/23/2009 11:02:49 AM, beetle wrote:
On 10/22/2009 5:09:30 PM, eden_mei wrote:
>On 10/22/2009 4:17:30 PM, beetle wrote:
>>So far all you have done is send data
>>buried in (horribly
>>rendered) mathcad binary worksheets --
>>i.e. data that cannot be
>>uploaded and tested by other Ms.
>
>
>While I would rarely defend jmg and his
>antics, it seems a bit disingenuous to
>claim that his data is inaccessible.
>Select any column, copy selection, and
>paste where you want.
>
>TTFN,
>Eden

I tried what I could to extract the data. I was unable to copy and
paste columns, only individual entires and like I said I am not
going to do that 2 X 600 times. This guy has the data in a
format that can be provided but chooses to be obstructive. If he
believes that the other Ms can't do as well as mathcad he'd
actually be more than happy to provide the data I would have
thought, instead he's being disingenuous. ...but further reading
it looks like he is the forum kook.

The data and model looked like a step decay process (like a
laser flash experiment in chemistry) which is trivial to model.

Got me ma popcorn and 3D glasses and the best seat in the House. Yessirree! Dangd ifn I ain't looking forrard to this ...

Stuart

As I'm getting bored watching the two of you standing at either end of the street, hands near pistols, waiting for the other to break, I've attached a worksheet that allows each array to be saved in csv format (could have been done by individual file components, but I'm too lazy to do that and I've only got M13 handy).

... just noticed that Eden and Richard have suggested other ways of getting the data too.

ptc-1368288 · ‎Oct 23, 2009

>This guy has the data in a format that can be provided but chooses to be obstructive.<<br> ___________________________

In over 80000 collaboratory exchanges in this forum, nobody has ever been obstructive. You simply speak about Mathcad that you know nothing about if you can't copy a data table and paste in Excel. Here is attached.

jmG

RichardJ · ‎Oct 23, 2009

On 10/23/2009 1:13:07 PM, jmG wrote:

>In over 80000 collaboratory
>exchanges in this forum,
>nobody has ever been
>obstructive.

Really? I know I have had that feeling of talking to a brick wall on a number of occasions (more than I care to count in fact).

Richard

ptc-1368288 · ‎Oct 23, 2009

On 10/22/2009 4:20:57 PM, beetle wrote:
...
The way to test ... about Mathcad vs other Ms for
curve fitting is to supply a data set and a model.
.........................

So far all you have done is send data buried in (horribly rendered) Mathcad binary worksheets -- i.e. data that cannot be uploaded and tested by other Ms.
________________________

At least one collab replied the data set was correct in my transmission.
Maybe your Mathcad and/or your server is/are all zombie. Download some work sheet from the PTC library, up until you get a data table and check if you have the same problem, check with your server. Here is the same work sheet again. Specify your version and kindly ask if some collab are willing to spare 5 seconds to check the attached works properly. Many of us collabs have that "Scientist/Engineering" expertise to make Mathcad work for them. What could make your Mathcad so sick ? If you just purchased, send it back for a new one, quite a rare case, never reported in 10 years (in this collab).

jmG

PS: Mathcad reads binary data, not the other 3M's ?

TheodoreM.Bones · ‎Oct 23, 2009

Yes, it works in MC11 from MC14MO.

ptc-1368288 · ‎Oct 23, 2009

On 10/23/2009 12:59:09 AM, study wrote:
>Yes, it works in MC11 from
>MC14MO.
___________________________

Thanks Theodore.

So, the problem is on the visitor's side ...
maybe not familiar enough with Mathcad ?

jmG

Ninetrees · ‎Oct 23, 2009

All I did was click on the upper left corner of the data table (M14), which selected the whole table. Then a ctrl-c and a ctrl-p into my text editor (similar to how it might be done in a spreadsheet), and I had 601 pairs of data. Perhaps the new visitor would have had better cooperation had he not introduced himself to the forum with such arrogance and rudeness. I would be interested in seeing comparisons of the products mentioned with tests that are more sophisticated than what I can come up with.

So, "beetle", let's put the shoe on the other foot. //You// provide the tests in the other Ms and the test data. Then perhaps we'll see how Mathcad fares. But I'm guessing that at this point you have offended so many folks here that you may not get the response you seem to have hoped for.

Rich
http://www.downeastengineering.com/

mzeftel · ‎Oct 23, 2009

Can we keep this discussion civil and refrain from further personal attacks?

I just read through this long thread and saw more than one post that was very borderline.

There used to be a site that ran comparisons of the 4 M's, but I don't remember the URL. Perhaps it no longer exists.

Mona

TomGutman · ‎Oct 22, 2009

Mathcad can write csv files. Or Excel files.
__________________
� � � � Tom Gutman

ptc-1368288 · ‎Oct 21, 2009

>Can you clowns come up with a ... curve fitting problem...<<br> ___________________________

I have added one (yellow, this one has no model function. Once you will have finished, I will post the graph for each function and the last one (yellow) as a fit in order for you to compare the Mathcad way of fitting data sets.

jmG

GuyB · ‎Oct 24, 2009

While this exercise might be mildly amusing, I'm not sure what the point of it is.

All the packages allow one to define curve fitting functions - if the package doesn't do something by itself we can program them to do what we need.

I'm not familiar with Maple, but I believe each package can do arbitrary curve fitting, including the incorporation of constraints, so it's really only a matter of style.

In the end, it will come down to which packages seem easier to use ... which will naturally be skewed in favor of what is most familiar.

Let the games begin 🙂

- Guy

PhilipLeitch · ‎Oct 13, 2009

And just to labour the point - I got two emails. By Australian anti-spam law that's illegal. You can only send out one email. After that the person has to either "opt-in" or not get any more emails from you.

I don't know what the situation is in USA.

HOWEVER - although it's unethical my care factor is zero, and I'm interested in the guff they send through.

Philip
___________________
Nobody can hear you scream in Euclidean space.

SteenGroðe · ‎Oct 15, 2009

On 10/13/2009 2:43:02 AM, pleitch wrote:
>And just to labour the point -
>I got two emails. By
>Australian anti-spam law
>that's illegal. You can only
>send out one email. After
>that the person has to either
>"opt-in" or not get any more
>emails from you.

In Denmark even one is illegal...

Steen Gro�e