getCyclopsProfileLogLikelihood {Cyclops} | R Documentation |
Profile likelihood for Cyclops model parameters
Description
getCyclopsProfileLogLikelihood
evaluates the profile likelihood at a grid of parameter values.
Usage
getCyclopsProfileLogLikelihood(
object,
parm,
x = NULL,
bounds = NULL,
tolerance = 0.001,
initialGridSize = 10,
maxResets = 10,
includePenalty = TRUE,
optimalWarmStart = TRUE
)
Arguments
object |
Fitted Cyclops model object |
parm |
Specification of which parameter requires profiling, either a vector of numbers of covariateId names |
x |
Vector of values of the parameter |
bounds |
Pair of values to bound adaptive profiling |
tolerance |
Absolute tolerance allowed for adaptive profiling |
initialGridSize |
Initial grid size for adaptive profiling |
maxResets |
Maximum allowed number of recomputing the likelihood when coefficient drift is detected. |
includePenalty |
Logical: Include regularized covariate penalty in profile |
optimalWarmStart |
Logical: Use optimal warm-starting when parallelizing evaluations |
Value
A data frame containing the profile log likelihood. Returns NULL when the adaptive profiling fails to converge.
[Package Cyclops version 3.5.0 Index]