Solve Block

[tietjee]

I am working on a project where I will being several solve blocks. I have had good success using solve. If I have defined a function in a solve block, can I use it in another solve block? I am pretty sure that you cannot have a solve block within another solve block. But MathCAD may treat a function differently.

Thanks for the support.

David Tietje

[RichardJ]

Yes. If you define a solve block so that it is a function, then that function can be used in another solve block.

Richard

[tietjee]

Thanks, I will give it a try

David Tg

[ptc-1368288]

>I am pretty sure that you cannot have a solve block within another solve block. <<br> _______________________

Will surely depend upon the structure of the solver(s) involved as well as the Mathcad versions(s). PTC seem to have played a lot in those blocks since Mathcad 11.2a and previous versions.

Attach the block as an image to be reconstructed by users of different versions.

jmG

[mzeftel]

You can not insert a solve block inside another solve block.

Mona

[ValeryOchkov]

On 1/22/2010 11:28:48 AM, MonaZ wrote:
>You can not insert a solve
>block inside another solve
>block.
>
>Mona

Can we insert if-operators in solve block?

See one task:
http://twt.mpei.ac.ru/MCS/Worksheets/chain-eng.xmcd


Val
http://twt.mpei.ac.ru/ochkov/v_ochkov.htm

[ptc-1368288]

On 1/22/2010 11:28:48 AM, MonaZ wrote:
>You can not insert a solve
>block inside another solve
>block.
>
>Mona
___________________________

I will try in 11, but the case would be unique.

Jean

[ptc-1368288]

On 1/22/2010 11:28:48 AM, MonaZ wrote:
>You can not insert a solve inside
>another solve block. [Mona]

On 1/22/2010 12:59:25 PM, jmG wrote:
>I will try in 11, but the case would be unique.[Jean]
______________________

1. You can insert root within root
2. Also root within Given/Find

Root-root illustrated w/o initials

Jean

[ptc-1368288]

On 1/22/2010 11:28:48 AM, MonaZ wrote:
>You can not insert a solve inside
>another solve block. [Mona]
_____________________________________

I will leave it here for 11 users to reconstruct as they may be interested, and for 14 users also ... Philip replied yesterday that the local assignment worked in 14 and that I should use 14 instead of 11, then a good test.

Note that at the white arrow pointing
Q<-- root(sol(u,a,b,c,d,n)-u,u)
The solver is root within root.



For the pros/cons, please read the text in graph.

Jean

[ptc-1368288]

... here is the work sheet for 11 users.

Does it work for the 14 users ?

and the note for those confused between "scalar" and "discrete"

>Disambiguation & Tip ��� nn, is the range variable [the argument] it is a scalar in Mathcad words and error messages. When pressing = it shows no row/col as well as Q(nn). They are display only @ the kernel and QuickPlot levels. Note that is displayed nested.
None of the 3 is an export, i.e: can't be spooled to file ... only the last value of the range would be
WRITEPRN , a well as Q(nn) & augment(nn,Q(NN)). Scalar is kernel volatile.
That the project is pregnant with 'nn', it is only slightly pregnant ! and consumed no hay [memory].
Scalar maths is a fairy: you see it but is not there ! But it produces off springs [magic].
On the other hand in discretized mode are created and resident @ the user level. They are results and as such can be spooled to file WRITEPRN, READPRN, or otherwise copied and pasted to other recipient.<<br>
That work sheet deserves the PTC monthly news. In few steps of a past collab application, it reveals lot more about the Mathcad structure than words. It is vital for new Mathcaders to distinguish the two form of executable maths, i.e: scalar vs discrete.

jmG

[ptc-1368288]

... read more:

In this work sheet just posted in the above thread, the Maple root solver is used. The same procedure applies in Mathematica, as you can see it fails on the low range. The Mathematica warning is about like this "when solving the InverseFunction is used, the solution might fail or not all solutions ..." . This is why their celebrated function has no inverse because of the numerical instability. However the same function has been inversed in Mathacad via the seeded root solver. In CAS, there are so many nuances of importance that knowing too much is sometimes not enough. All "InverseFunction" have asymptotic region(s), that's where any solver fails. You can't solve for x on an horizontal line ...

jmG