Calculate gradient timings on a model for different metrics
Usage
benchmark_metrics(obj, times = 1000, metrics = NULL, model_name = NULL)
Arguments
- obj
A TMB object
- times
How many evaluations to do
- metrics
A character vector of different metrics to benchmark
- model_name
An optional character name for the model, if
NULL will pull from the DLL name
Value
A data.frame containing the median gradient timing
(time), the percent sparsity of \(Q\) and the dimension of
the model (npar).