[FW Beta3] Solve APP failed... Can't understand...

[dg1969]

Hi,
Here is an example of a system equations...
degree mode is on...

Code:


E1: R+T-P*SIN(-15)=0
E2: S-Q+P*COS(15)=0
E3: -A*T+P*(D*SIN(-15)-C*COS(15))+B*Q=0

Unknowns are R, S and Q

So in fact the 1rst eq give R and then we have S and Q with E2 and E3

Code:


A:=2.62
B:=25.88
C:=0.97
D:=1.84
T:=450
P:=-1400

If I check only E2 and E3 The solutions for S and Q are OK:


But if I try to solve the three equations in a single system (E1, E2 and E3 are check) and with the num view complete like these:

The solve app change T and P !!! But T and P are not checked !! ? as you can see :

R solution is OK but S and Q are not !!!
The approx solutions are:
R=-88 ; S=1383 and Q=-31

Any help are welcome...
I do not understand what I'm doing wrong...

[Carlos295pz]

It seems to be just a problem of insertion of values, the calculations are correct. The App is inserting the solutions in the first boxes, it is not respecting the position of the selected boxes.

[Carsen]

I don't know what this unusual phenomenon is but it worked with my HP Prime perfectly. Every value is with its correct number.

[Tim Wessman]

Thanks for providing the files. Makes it easier to investigate.

[dg1969]

(01-06-2018 08:44 PM)Carlos295pz Wrote:  It seems to be just a problem of insertion of values, the calculations are correct. The App is inserting the solutions in the first boxes, it is not respecting the position of the selected boxes.

Yes indeed, well seen. The problem is now in very good hands. Potentially there is no more problem.

[Helge Gabert]

Are you sure that S should equal approx. 1383?

With fsolve, I get approx. 1321 for S., with R and Q as you indicated above.

[roadrunner]

In the OP's code E2 is indicated as:

E2: S-Q+P*COS(15)=0

But the screen shot shows E2 as:

E2: S+Q+P*COS(15)=0

Could this be part of the problem?

-road

[Tim Wessman]

Thank you for catching this. Indeed, the solution storage into the checked variables wasn't handling gaps in the checkmarks. This is now corrected!

[dg1969]

(01-08-2018 05:15 PM)Tim Wessman Wrote:  Thank you for catching this. Indeed, the solution storage into the checked variables wasn't handling gaps in the checkmarks. This is now corrected!

Thank you Tim !