gsal said:
That's funny, Nemo...it looks to me that you actually made my point, although the "tone" of you writing might have tried to advocate against it. Or I am missing something.
I think so.
First, there is no such thing as a systematic error being introduced into two different systems...
There most certainly is. I've done it myself and I've encountered in many times. If I've misread the original text or operated from a given set of incorrect assumptions, then all I will get by implementing my interpretation in n different languages is n + k (
where k = number of debugs) different wrong results (and maybe not even the same final wrong result!)
that's the good thing about being force to type the same formulas into two different systems...if you know what you are doing, with the formulas AND the system...then you should get it right on both.
Your 'if' and 'should' are noted.
Also, it is funny that you said that "there is no real substitute for looking at the [mathcad] worksheet" when right before you indicated several traps with some similar looking equations with different meaning...
In what way is it funny? I have considerable experience of looking at other people's worksheets and many errors arise from inadvertent use of literal v indicial subscripts or implied multiplication v function calls. They are the kind of thing that new users haven't yet gained familiarity with (or don't yet know) and for experienced users they form the equivalent of the 'missing semi-colon' so beloved of other languages.
...so, to that end...there is nothing like good old plain text looking equations like the ones you would need to put together with python, for example.
The problem with good old plain text, and the major reason I avoid it like the plague, is that it is often verbose and long expressions get difficult to write correctly and read correctly. There is a reason Mathematics uses more concise notation than FORTRAN and don't get me started on Excel! One of the reasons I use Mathcad is that it is a lot easier to write and verify the implementation against the specification and easier for non-programmers to validate the implementation - summation, integral and factorial operators look familiar and are easier to interpret than code, particularly when the results are presented right alongside.
Anyway...now for the now for the one piece of advice I always give: "Baby steps...baby steps".
No disagreement from me there.
rottbull: I would recommend you start typing your equations in a different piece of software; you don't need to type EVERYTHING...just one at a time until you find where your answers start diverging between the two systems...that where your problem will be.
I would recommend taking those baby steps in Mathcad in the first instance. Break the problem down into smaller steps (indeed, in many cases, this is just a simple matter of evaluating the worksheet at each stage) and check the results make sense.
I would only resort to another language when getting really bogged down; I learned Python, J, C, C++, C#, FORTRAN, multiple dialects of Basic, Pascal, Modula, ADA, FORTH, LISP,PROLOG, Assembler and goodness knows what else by working
within the languages.
Still, different horse are suited for different courses and it may be that gsal's method works better for you.
NM
