Maximum likelihood estimation and analysis with the bbmle package
Benjamin M. Bolker · 2011
4 Technical details 17 4.1 Profiling and confidence intervals . . . . . . . . . . . . . . . . . . 17 4.1.1 Estimating standard error . . . . . . . . . . . . . . . . . . 17 4.1.2 Profiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 4.1.3 Confidence intervals . . . . . . . . . . . . . . . . . . . . . 19 4.1.4 Profile plotting . . . . . . . . . . . . . . . . . . . . . . . . 20 The bbmle package, designed to simplify maximum likelihood estimation and analysis in R, extends and modifies the mle function and class in the stats4 package that comes with R by default. mle is in turn a wrapper around the optim function in base R. The maximum-likelihood-estimation function and class in bbmle are both called mle2, to avoid confusion and conflict with the original functions in the stats4 package. The major differences between mle and mle2 are: