CalculationCenter 7

[mrMikee]

As I looked over posts in the spreadsheet forum, I started to wonder why more people don't use a program like CalculationCenter (CC2) from Wolfram Research. I admit I am not very experienced with Excel and have not used it much. Most of my calcualtions over the years has been with BASIC and VisualBASIC, but during the last year I am using CC2 more and VB less.

If anyone is interested in this topic please respond. I am curious about what others have experienced with calculations and various software answers. In other words, an overview, some personal opinions, perspectives, and so on.

A brief background on CC2. It seems to be a lite version of mathematica but with some special commands, and costs only $295. Mathematica, Maple, and Matlab are very expensive, especially for people like me who have to buy their own software. Even Mathcad is getting expensive at $1200.

I guess what I'm saying is that I like CC2, but would like to know what other engineers think.

-Mike

 

[GregLocock]

For those of you looking for a math language, try Scilab from

http://www.scilab.org

It is very similar to Matlab, except that it is free. It is remarkably good. There are many toolboxes available for it, including an FEA package.

To be honest I found the learning curve for Matlab was a great deal quicker than Mathematica, which I never really got to grips with.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[mrMikee]

GregLocock,

Free is good.

I have just downloaded it to do some experimenting.

Thanks,
-Mike

 

[Gearhead]

Hi Guys

What about TK Solver 5 and Interactive Roark's for you structural folks.
TK five also has the ability to use EXCEL AS THE INTERFACE.
You can do engineering calculations in TK and package a free app with the TK Player. Check out

http://WWW.uts.com

for a piece of software that has been around since before all the rest yet has been improved with excel users in mind.
Never heard of it? No suprise they never advertize!!!

 

[cnuk]

Good to see someone else mention TKSolver! I am an Excel junkie, but recently bought TKSolver v5 and use the Tk-Excel Toolkit all the time. Develop the mathematical model in TKSolver which is as simple as writing down the equations. Fill in the variables you know and TK will solve for the remaining unknowns (providing you have given it enough to go on). Great for backsolving and trying "what-if" scenarios that is just painful in Excel. Once you have the TK model working, use the TK-Excel Toolkit to make a spreadsheet that has built in units conversions, etc....really nice. They have taken it one step further now by providing a TK-Excel player so users without TKSolver can display and edit the TK-Excel models you have created. Oh yeah, and don't forget all the built in functions to lookup properties of common fluids....very slick for fluids and heat transfer calculations! Tech support is great.

No I don't work for UTS...just a happy user.

 

[JayMaechtlen]

I've used Mathcad occasionally for many years. Nice tool, much more elegant than Excel for lots of applications. (as someone else commented, complex Excel equations tend to be unreadable and difficult to uderstand/maintain)
Fortunately or unfortunately, my job never quite required enough math for me to get proficient at Mathcad.
Scilab sounds pretty neat, but I'll probably just stick with Excel and Mathcad 2k for now.
Wasn't TK Solver the pioneer about 20 years ago?
Regards
Jay

Jay Maechtlen

http://home.covad.net/~jmaechtlen/

 

[HgTX]

Is there any difference between CC2 and MathCAD?

In my brief tour through the world of design (I'm usually in fabrication), the office I was in DIDN'T HAVE MATHCAD. I thought that that was almost as bad as not letting us have pencils and paper. People used a lot of Excel worksheets, but if you wanted to show your work, you did it longhand.

I wound up sort of faking MathCAD by extensive use of the STRING function in Excel. It tripled my time to set up a worksheet, but at least I could show my work in a legible and easily modifiable manner.

Ridiculous.

Hg

Eng-Tips guidelines: faq731-376

 

[GregLocock]

http://www.scientific-computing.com/scwsepoct02reviewfriendlymaths.html

gives an extremely brief overview of CC2.

I can't get a picture of whether you can write a 10 page live mathematical report inside CC2, which you can in Mathcad.

Frankly I haven't seen anything else that even comes close to MathCAD when I'm exploring new concepts, but that looks as though it would do it.

However, in my current job I am using Matlab more and more, partly because they won't spring for MathCAD, and partly because we already have a lot of Matlab. I think Excel is inappropriate in an ISO9000 organisation as well.


Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[HgTX]

Why is Excel inappropriate in ISO9000?

Hg

Eng-Tips guidelines: faq731-376

 

[mrMikee]

HgTX,

I haven't used MathCAD but what I have seen suggests that MathCAD produces a very nice looking report. The CC2 print out is more plain looking but fine for me. The longest notebook I worked on was about 20 pages long and it would run very fast. My son works with MathCAD but I don't really know the two compare. My guess is that both would work out well for most engineers doing everyday engineering things, but if you are going to do a PhD or work in a research facility, you would be better off with Maple, Mathematica, or Matlab.

The following links compare the 'big' systems.

http://amath.colorado.edu/computing/mmm/index.html

http://amath.colorado.edu/computing/mmm/syntaces.html

Note when looking through these links that the syntax in CC2 is generally the same as Mathematica except it has a smaller set of commands and some of the commands have been replace with the CC2 equivalent.

Anyone interested should download the 15 day free trial to make sure it's something they want. For someone like me with many years of BASIC programing it was a real shift in thinking from what I was used to. And I didn't like it at first.

-Mike

 

[GregLocock]

HG

1) version control

2) How do you know that someone hasn't changed the result of a safety related calculation by inserting a cell with the following pseudologic

ans=if (and ((ans>x),( ans<x*1.0000001)),wronganswer,ans)

Where wronganswer causes your bridge to fall down. No raesonable amount of error checking by comparing handcalcs with output could catch that, so in theory you need to check the formula in EVERY cell.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[HgTX]

Version control is a problem with just about anything.
The "black box" nature of the spreadsheet formulae is more of a concern.
In theory anything that generates ascii data files should be suspect as well, no, since those can be tampered with? Or is outright crime not part of the ISO consideration?

Hg

Eng-Tips guidelines: faq731-376

 

[IRstuff]

As you might recall from other discussions herein, ISO9000 makes no value judgements on the validity of your process.

If your process calls for lying and cheating, and it's documented, and you follow it; you, too, can be ISO9000 certified. Therefore, any criminal operation can also be ISO certified. [Courtesy of Dewey, Cheatham & Howe]

TTFN

 

[GregLocock]

(Laughing) um, none of our ISO procedures actually talks about engineers making mistakes, let alone deliberately falsifying results.

We do 'error check' by building and testing prototypes, a luxury you don't have.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[btrueblood]

IRstuff,

Man you've just created a nightmare - the Mafia is going to go ISO 9000 now. I can just see the future mugging victim being handed a "customer feedback survey" to fill out. Maybe with an incentive being a discount on his next mugging?

Actually, that last idea isn't mine, thanks Mr. Pratchett.

 

[GregLocock]

Oh, and version control. If I have a text based program, such as Matlab, or a conventional programming language, then I can, if I wish, check for changes by using WinMerge or whatever on the source code, and then see if the changes look OK (not foolproof, but at least feasible).

How can I do that for a 20 Mb spreadsheet?

Actually there are a couple of ways to improve the integrity of a ss, but they aren't even moderately foolproof.

a) use the ss for data entry only and use VBA for everything computational.

b) write protect everything except data entry



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[IRstuff]

Oh, you mean like cashback program? ;-)

Greg,

You issued a challenge and here is a possible solution. If you select display formulas mode and save as .CSV, you can open the .CSV file in Word and do a version comparison.

Brute force and clearly inelegant, but presumably doable, although I'll confess that I've not tried this on a 20MB file.

TTFN

 

[mrMikee]

Or you could open the .CSV with a VB program which would be able to scan a 20MB relatively fast. I assume you only want to know if the files are the same or not.

It's not pretty but it should work. Sometimes you can paint yourself into a corner.

I have found that when writing programs, you don't always know where you are going until you get there.

At least for me.

-Mike

 

[GregLocock]

Man, that "display formulas' mode is ugly. Very useful though, thanks.

I agree, a Winmerge report on the csv of that display would satisfy my theoretical desire to make sure that the code is unchanged.






Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[jghrist]

I found the following VBA function definition to be very helpful in documenting spreadsheets:

'Function to display cell formula
'copied from Eng-Tips thread770-37797 12/17/02 electricpete post

Function GetFormula(Cell)
GetFormula = Cell.Formula
End Function

I have it in a "General" spreadsheet that I use as a template to start all of my spreadsheets. "General" also has separate Introduction, Input, Model, and Report sheets to aid in keeping things understandable and documented.

 

[electricpete]

I learned about that getformula from this forum and I really like it as well.

There is one small hiccup, if you are using named cells (vs just cell address), the name won't show up in the getformula display unless the name was created before you created the formula. Sometimes I end up having to go back and re-enter formula's just to capture a later-created name in the getformula display.

But when I have to work in excel, getformula defintely helps make it less user un-friendly.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.