bbmle

bbmle copied to clipboard

Not clear whether this should be merged, but we wanted you to see. I was not able to easily reproduce any problems without underflow control code, so Mike is going...

dushoff

Something like ``` mu_expr

bbolker

``` library(bbmle) y

parksw3

for example, consider ``` library(bbmle) set.seed(101) z = rpois(100,lambda=5) m1 = mle2(z~dpois(lambda=L),start=list(L=4),data=data.frame(z)) qAICc(m1,nobs=100,dispersion=1.2) qAICc(m1,m1,nobs=100,dispersion=1.2) ## !! ```

bbolker

When profiling following an mle2() fit, profiling often finds a new better MLE. Currently profile() then stops and returns the new MLE. This is useful, but it seems like it...

marmkilpatrick

Support confidence and prediction intervals in predict method?

3

For a specific application I would like to be able to calculate confidence and prediction intervals. Would it be hard to incorporate these in the predict method of bbmle?

tomwenseleers

implement continuation methods for profiles

1

as far as I can tell, profiling is using the same starting values as the original call. This is inefficient and particularly bad when the likelihood surface is sufficiently weird...

bbolker

provide analytical gradients for formula-based methods

3

bbolker