NonlinearModelFit ... please find me only integers
Ktota
how can i tell NonlinearModelFit that a parameter it is supposed to
optimize should be only an Integer?
thank you
Konstantin
Bill Rowe
>how can i tell NonlinearModelFit that a parameter it is supposed to
>optimize should be only an Integer?
As far as I know, this cannot be done using NonlinearModelFit.
NonlinearModelFit seems to be designed to do non-linear
regression with the usual assumptions. This is an unconstrained
optimization problem. Once you impose a constraint on any
parameter, the problem becomes a constrained optimization
problem. See:
tutorial/UnconstrainedOptimizationOverview
tutorial/ConstrainedOptimizationOverview
=46indFit or NMimimize can be used to find the best fit parameters
with constraints. Look at the second entry in the online
documentation for FindFit for syntax details.
Bob Hanlon
You can try using Round in your model
data = {{0, 1}, {1, 0}, {3, 2},
{5, 4}, {6, 4}, {7, 5}};
f[x_] = NonlinearModelFit[data,
Log[a + b x^2], {a, b}, x] // Normal
Log[1.4263297975161129*x^2 + 1.5063204891556876]
f2[x_] = NonlinearModelFit[data,
Log[a + Round[b] x^2], {a, b}, x] // Normal
Log[x^2 + 1.5806558441280811]
f3[x_] = NonlinearModelFit[data,
Log[Round[a] + b x^2], {a, b}, x] // Normal
Log[1.444296658409274*x^2 + 1]
f4[x_] = NonlinearModelFit[data,
Log[Round[a] + Round[b] x^2], {a, b}, x] //
Normal // Quiet
Log[x^2 + 1]
Plot[Evaluate[Tooltip[#[x], #] & /@
{f, f2, f3, f4}], {x, 0, 7},
Epilog -> {Red, AbsolutePointSize[4],
Point[data]},
Frame -> True,
Axes -> False,
PlotRange ->
{{-0.25, 7.25}, {-0.25, 5.25}},
ImageSize -> 600]
Bob Hanlon
---- Ktota <nuk...@gmail.com> wrote:
=============
Hi there,
how can i tell NonlinearModelFit that a parameter it is supposed to
optimize should be only an Integer?
thank you
Konstantin