A data frame of precomputed [optim_sample()] results across a factorial grid of design conditions for the Joint Hierarchical Model. Each row contains one parameter configuration and its corresponding optimization result.
Use [available_configs()] to view the grid of available conditions, and [get_sspLNIRT()] to retrieve results by parameter values.
Format
A data frame with two list-columns (in this order):
- cfg
List of `"sspLNIRT.design"` objects, each containing the full set of input parameters used for the optimization (e.g., `thresh`, `out.par`, `K`, `mu.item`, `cov.m.person`, `meanlog.sigma2`, `seed`, etc.).
- res
List of `"sspLNIRT"` results, each as returned by [optim_sample()]. Contains `N.min`, `res.best`, `comp.rmse`, and `trace`.
The design grid varies the following parameters: - `out.par`: `"alpha"`, `"beta"`, `"phi"`, `"lambda"` - `thresh`: RMSE thresholds (e.g., 0.05, 0.10, 0.15, 0.20) - `K`: test length (e.g., 30, 50) - `mu.alpha`: mean discrimination (e.g., 0.6, 0.8, 1.0, 1.2, 1.4) - `meanlog.sigma2`: residual variance level (e.g., log(0.2), log(0.4), log(0.6), log(0.8), log(1.0)) - `rho`: person parameter correlation (e.g., 0.2, 0.4, 0.6)