lgj200306 writes:
>
> Thanks Scott, Matteo and Mollie very much.
> My function is a little complicated, and composed by several sub-functions,
> so I did not attach it in the
> previcous mail. I will check and try to simplify it, and
> then try more initial values.
> Thanks very much for your
Thanks Scott, Matteo and Mollie very much.
My function is a little complicated, and composed by several sub-functions, so
I did not attach it in the previcous mail. I will check and try to simplify
it, and then try more initial values.
Thanks very much for your attention and best wishes for you!
Hi,
Yes, you should probably choose different initial values. You might be dividing
by zero or taking the log of a negative number or zero. If the initial values
really seem sensible, then check the code of your function for mistakes.
Without seeing your function, and call to mle2, it's difficul
HI Guojun,
I struggled a little too with mle2, and had the same error message.
It could be that the initial value you insert do not satisfy the assumption of
the function you are trying to fit.
It may be helpful if you say something more on it.
Cheers
matteo.
Il giorno 7 Nov 2011, alle ore 00:5
Hi, all
I want to estimate four parameters using function mle2 (package bbmle). An
error produced as follows:
>Error in optim(par = c(10, 0.01, 0.1, 0), fn = function (p) :
>initial value in 'vmmin' is not finite
At first,I thought it might be the problem of improper initial values. Then I
ch
