[scilab-Users] lsqrsolve

[vinsz]

On 31/03/2010 11:18, Maso Ricci wrote:
> Hi,
>
> I am using the lsqrsolve function to fit some data.
> I wonder whether lsqrsolve accepts contrains in order to force the 
> solution to be positive.
>
> thanks in advance for any comments and suggestions

A common trick to force a parameter to remain positive is to use the 
exp() function in the solver, for instance:

function er = to_fit(X, m)
   a = exp( X(1) )
   b = X(2)
      [...]
endfunction

x = lsqrsolve([log(a0), b0], to_fit, N)
a_fit = exp(x(1))

hope it helps,
Best regards,