error in Solve app .__.

[nakjaen]

Guys, i dont understand why this application give me error

I introduce a system of equations, with variables defined, and i go to "num" and then press "solve" and give me a error

Ex:




I can solve whit matrix, but i prefer use Solve app, how can i get the solution?
Its necessary what i input a value near with the solution?

[jebem]

Before pressing "Solve", try with different seed values using the "Edit" option, for instance:
IC: 5
S: 9
The math experts here may help to explain why the Prime is not able to do what wolffram alpha can do at first try without having to specify successive seed values until one gets an answer.

[nakjaen]

But its imposible to know try with differents values in some equations
i input
IC: 3
S: 4

and say error again, its a problem for me -___- , i can't find out the seed values

only have a resolution with "solutions values" no seed values.

[jebem]

(11-02-2014 04:07 PM)nakjaen Wrote:  But its imposible to know try with differents values in some equations
i input
IC: 3
S: 4

and say error again, its a problem for me -___- , i can't find out the seed values

only have a resolution with "solutions values" no seed values.

Did you try with my seeds?

[nakjaen]

Yes, but i need a lot of time to have the values.....

[jebem]

(11-02-2014 04:14 PM)nakjaen Wrote:  Yes, but i need a lot of time to have the values.....

The Solve app works as advertized in the User Guide...
You supply a seed value to one variable, and the app tries to solve for the other variables based on your seed.

Again, let us wait and see if the math experts can give an explanation for this behavior.

[akmon]

fsolve command solves this system with no guess numbers needed.
IC=3,76
S=18,82

The more I use the calculator, the more I use fsolve.

[Tim Wessman]

(11-02-2014 10:33 PM)akmon Wrote:  The more I use the calculator, the more I use fsolve.

That is actually what it is doing internally, however it provides the guesses you gave it. In this case a guess of 0 really messes it up.

[akmon]

Oh, hell, you are right. As I solved it previously in the solve app, fsolve command found the solution inmediatly.
I remember the program solvesys, on hp48 and 49 from sune bredal. What kind of algorithm did he use, that never failed to me?