Profiling a Log-Likelihood to Calculate Confidence Intervals v1.1.1

Latest

paulnorthrop released this 27 Jan 00:17

· 2 commits to main since this release

a5b2911

profileCI 1.1.1

When the argument epsilon to profileCI() is negative, monotonic cubic spline interpolation is now used instead of quadratic interpolation. Quadratic interpolation could fail in some cases.
In the documentation of the argument object to profileCI() it is emphasized that the parameters in the fitted model object must have names.
Fixed bugs in profileCI-internal.R to avoid storing some slightly incorrect parameter values near the confidence limits in the attribute "for_plot" in the returned object.

Assets 2