Hello,

Here is the example of one of my working files presenting the solving of eleven nonlinear equations with eleven unknowns.
File - Primer38a.sm
Picture :


It seems that the standard Newton-Raphson will make the job done here, and I am quite satisfied for that. Unfortunately, the roots() function will not be useful for this example. Actually, I do not know if this could even be solved by roots() because the calculation took quite a long time - I gave it up after 10 min.

The solution presented here is not so elegant but I hope that it would be useful to someone. On the other hand, this is another of my "cry for help" in order to potentially motivate some good willing people to extend the SMath with some more powerful numerical procedures. SMath deserves that in any sense.

Regards,
Radovan

Hello,

Unfortunately, this will not work in v.0.92 . It seems I have to give it up from this example at the moment.

Regards,
Radovan

Why? It works well for me.

Regards.

Wrote

Why? It works well for me.

Strange

This is is v0.92 - calculation still not finished.


This is the same file in 0.91



Have no idea what has going wrong. I'll try tomorrow on some other computers - it has already happened that some things did not work on my home computers and worked for you and some other users. Hmm... I am not quite happy about it .

Regards,
Radovan

EDIT: The same on WinXP and Ubuntu

Hi,

Unfortunaltely same result for me using Radovan's file on v0.92.

Regards

Steve

I figured out what has going on...Huh

Look at the picture please:


The constant "e" is not known inside the line() environment (v0.92). I used in the previous file exp() instead and it works now . I just looked at the " - not defined" error and asked myself what could not be defined - did not have a clue because it did not tell me what has not been defined. By inspecting step by step, I get the error in one step "e - not defined" and figured out what might be the reason. I am not quite sure why the calculation did not stop here. Maybe it continued in the symbolic fashion - who knows for how long. I hope I am right about it.

Another problem is that there is not known "pi" - п now. This might be the cause of error in calculating degrees directly. We can not use п as a constant inside line() as well. I also suppose that this comes from the changing of imaginary unit "i" in v0.92 and suppressing it as a function argument in the function definition. Btw, "i" as a constant can be used in inside the line() environment, contrary to e and п.

Andrey, could you check this out please.

Regards,
Radovan

The same as here: http://en.smath.info/forum/default.aspx?g=posts&t=1217