Displaying 1 result from an estimated 1 matches for "mtilt".
Did you mean:
milt
2012 Apr 14
0
R-help: Censoring data (actually an optim issue
...TRUE TRUE
[1] 7.138974e+02 9.931285e-04
Save results from method BFGS
> zz
method par fvalues fns grs hes rs conv KKT1 KKT2
2 BFGS 0.2832468, 52.6096161 100.8794 10 1 0 -2 3 TRUE TRUE
1 NM 0.2827563, 54.3551491 100.8779 53 0 0 1 0 TRUE TRUE
mtilt xtimes meths
2 NA 0.11 BFGS
1 0.0005130808 0.57 NM
There were 50 or more warnings (use warnings() to see the first 50)
>
Note the Hessian eigenvalues are pretty bad. And the warnings are the NaN's produced.
This is with just the 2 default methods. When I tried "all.me...