Description Usage Arguments Details Value Examples

This function allows to run a simulation study of mpt2irt models. Data are
generated either from the Boeckenholt Model (`genModel = "2012"`

) or
from the Acquiescence Model (`genModel = "ext"`

). Subsequently, one or
both of these models are fit to the generated data using either JAGS or Stan.
The results are saved in an RData file in `dir`

.

1 2 3 4 5 6 7 8 | ```
recovery_irtree(rrr = NULL, N = NULL, J = NULL, prop.rev = 0.5,
genModel = c("ext", "2012"), fitModel = c("ext", "2012", "pcm", "steps",
"shift", "ext2"), fitMethod = c("stan", "jags"), theta_vcov = NULL,
betas = NULL, beta_ARS_extreme = NULL, df = NULL, V = NULL, M = 500,
n.chains = 3, thin = 1, warmup = 500, method = "simple",
outFormat = NULL, startSmall = FALSE, df_vcov = 50, dir = NULL,
keep_mcmc = FALSE, savext_all = FALSE, savext_mcmc = TRUE,
add2varlist = c("deviance", "pd", "popt", "dic"), ...)
``` |

`rrr` |
Sequence of integers (e.g., |

`N` |
number of persons |

`J` |
number of items. Can be a vector for multiple traits (e.g., J=c(10,10,10)). |

`prop.rev` |
number of reversed items. Can be a vector for multiple traits(e.g., prop.rev=c(5,3,5)/10) |

`genModel` |
Character. The data generating model (either "2012" or "ext"). |

`fitModel` |
Character. The model for data analysis ("2012", "ext", or both as vector c("2012", "ext")). |

`fitMethod` |
Character. Whether to use "stan" or "jags". |

`theta_vcov` |
true covariance matrix of response processes (order: middle, extreme, (acquiescence), trait). standard is diag(3) / diag(4). Can be a vector of variances (not SDs). |

`betas` |
Optional list. May have entries |

`beta_ARS_extreme` |
Numeric. Only for |

`df` |
degrees of freedom for wishart prior on covariance of traits (default: number of processes + 1) |

`V` |
prior for wishart distribution (default: diagonal matrix) |

`M` |
number of MCMC samples (after warmup) |

`n.chains` |
number of MCMC chains (and number of CPUs used) |

`thin` |
thinning of MCMC samples |

`warmup` |
number of samples for warmup (in JAGS: 1/5 for adaptation, 4/5 for burnin) |

`method` |
Passed to |

`outFormat` |
either "mcmc.list" (can be analyzed with coda package) or "stan" or "runjags" |

`startSmall` |
Whether to use random starting values for beta sampled from "wide" (FALSE) or "narrow" priors (TRUE; beta and theta closer to 0; might solve problems with slow convergence of some chains for extreme starting values). |

`df_vcov` |
Numeric. Degrees of freedom for wishart distribution from which the variance-covariance matrix for generating the data is drawn. |

`dir` |
Path to directory where results should be stored, |

`keep_mcmc` |
Logical indicating wheter to retain, besides a summary of the parameters, the raw mcmc samples. |

`savext_all` |
Logical indicating wheter to save the output from Stan/JAGS in an external RData file. |

`savext_mcmc` |
Logical indicating wheter to save the mcmc samples in an external RData file. |

`add2varlist` |
Additional variables to monitor (e.g., |

`...` |
further arguments passed to |

Note that a text file "progress.txt" is written (and updated) to `dir`

informing you about the progress of the simulation.

Function does not directly return anything but saves an external
RData file to `dir`

. This object is a list containing the generated
parameters in `sim-results$param.sum$gen`

, fitted parameters and other model fit
information in `sim-results$param.sum$foo`

, as well as a summary of the setup.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 | ```
## Not run:
recovery_irtree(rrr = 1:2, N = 20, J = 10, genModel = "ext", fitModel = "ext",
fitMethod = "stan", M = 200, n.chains = 2, warmup = 200,
dir = "~/")
# run multiple simulations in parallel using the 'parallel' package
no_cores <- parallel::detectCores() - 1
cl <- parallel::makeCluster(no_cores)
parallel::clusterApplyLB(cl, x = 11:13, fun = recovery_irtree, cores = 1,
N = 20, J = 10, genModel = "ext", fitModel = "ext",
fitMethod = "stan", M = 200, n.chains = 2, warmup = 200,
dir = "~/")
parallel::stopCluster(cl = cl)
## End(Not run)
``` |

hplieninger/mpt2irt documentation built on Aug. 4, 2018, 10:52 a.m.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.