Hello,
I have a question about the optimizer tool in Multiple Run Fitter. I would like to vary one of the parameters as an integer. Please let me know if there is an way to do that.
Thank you in advance.
Yamin
Optimizing Integer Increment

 Site Admin
 Posts: 5784
 Joined: Wed May 18, 2005 4:50 pm
 Location: Yale University School of Medicine
 Contact:
Re: Optimizing Integer Increment
That's not going to work with Praxis. In general terms, optimization means finding the particular combination of values for parameters a_0, a_1, . . . a_n1 that minimizes the objective function f(a_0, a_1, . . . a_n1). Praxis requires that each a is continuous over some interval. If you know that one of the parameters, say parameter a_n1, can only have integer values, and you can furthermore stipulate that this parameter can have only one of m different values, then you could recast the problem like so:
Given a set of integer values i_0, i_1, . . . i_m1, define functions g_k(), where k = 0, 1, . . . m1, such that
g_k(a_0, a_1, . . . a_n2) = f(a_0, a_1, . . . a_n2, i_k)
Then you have m different subproblems: for each k, find the combination of values for a_0 . . . a_n2 that minimizes g_k. The final step is to find which g_k has the smallest minimum.
Otherwise you could give up on Praxis and add another optimizer of your own choice.
Given a set of integer values i_0, i_1, . . . i_m1, define functions g_k(), where k = 0, 1, . . . m1, such that
g_k(a_0, a_1, . . . a_n2) = f(a_0, a_1, . . . a_n2, i_k)
Then you have m different subproblems: for each k, find the combination of values for a_0 . . . a_n2 that minimizes g_k. The final step is to find which g_k has the smallest minimum.
Otherwise you could give up on Praxis and add another optimizer of your own choice.
Re: Optimizing Integer Increment
thank you so much. I wrote a loop and it works fine.