A general framework for statistical simulation, which allows researchers to make use of a wide range of simulation designs with minimal programming effort. The package provides functionality for drawing samples from a distribution or a finite population, for adding outliers and missing values, as well as for visualization of the simulation results. It follows a clear object-oriented design and supports parallel computing to increase computational performance.
Changes in simFrame version 0.5.3
+ Number of worker processes in examples for parallel computing is now limited to 2.
Changes in simFrame version 0.5.2
+ Bugfix in 'clusterRunSimulation' regarding export of internal object to worker processes. + Bugfix in demo for model-based simulation with parallel computing.
Changes in simFrame version 0.5.1
+ Slight optimizations in C++ code to compute inclusion probabilities. + Modified vignette according to changes in package 'robCompositions' (function 'invilr' renamed to 'isomLRinv'). + Moved vignettes to 'vignettes' folder.
Changes in simFrame version 0.5.0
+ Using package 'parallel' now for parallel computing instead of 'snow'. + 'runSimulation' and 'clusterRunSimulation' no longer produce a warning when simulations are performed separately on different domains.
Changes in simFrame version 0.4.4
+ Rewrote parts of the C++ code.
Changes in simFrame version 0.4.3
+ S4 generic 'getCall' now has '...' argument to be consistent with R 2.14.0 S3 generic. + Updated author affiliation.
Changes in simFrame version 0.4.2
+ Midzuno and Tille sampling now give a warning instead of throwing an error when all inclusion probabilities are outside the interval (eps, 1-eps).
Changes in simFrame version 0.4.1
+ Added column to example data 'eusilcP' that indicates the main income holder of each household.
Changes in simFrame version 0.4
+ Implemented 'runSimulation' and 'clusterRunSimulation' methods for mixed simulation designs. + Added control class "TwoStageControl" and corresponding 'setup' and 'clusterSetup' methods for two-stage sampling. + Probability weights or inclusion probabilities for unequal probability sampling can now be supplied by specifying a variable of the population data. + Using Rcpp now to incorporate C++ instead of C code. + Added functions 'brewer', 'midzuno' and 'tille' for Brewer, Midzuno and Tille sampling, respectively (corresponding to 'UPbrewer', 'UPmidzuno' and 'UPtille' from package 'sampling', but faster C++ implementation).
Changes in simFrame version 0.3.7
+ Added JSS paper to references in help files and vignettes.
Changes in simFrame version 0.3.6
+ Changes in package vignettes.
Changes in simFrame version 0.3.5
+ Improved code in 'CITATION' file. + Updated references in package vignettes.
Changes in simFrame version 0.3.4
+ If 'nrep' is not supplied in 'runSimulation' or 'clusterRunSimulation', the corresponding slot of the resulting "SimResults" objects now contains a zero-length numeric vector instead of 0.
Changes in simFrame version 0.3.3
+ Changes in help file for package.
Changes in simFrame version 0.3.2
+ Changes in help file for accessor and mutator functions.
Changes in simFrame version 0.3.1
+ Bugfix in 'clusterSetup' solved problem with inclusion probabilities for group sampling if the data are not sorted according to the groups.
Changes in simFrame version 0.3
+ Control class "NAControl" now allows to specify an auxiliary variable with probability weights for each target variable. + To improve computational performance in simulation studies, 'contaminate' and 'setNA' no longer check if auxiliary variable(s) with probability weights are numeric and contain only finite positive values ('sample' still throws an error in these cases). + Bugfix in 'setup' solved problem with inclusion probabilities for group sampling if the data are not sorted according to the groups. + Classes "SampleSetup" and "SimResults" now contain slots for the control object(s) used. Accordingly, some redundant slots were removed from class "SampleSetup". + Added package 'sampling' to 'Suggests' section of 'DESCRIPTION' file, minimal changes in 'simDensityplot', 'simXyplot' and in help file 'tail-methods' so that 'R CMD check' and 'R CMD INSTALL' no longer produce warnings or notes.
Changes in simFrame version 0.2
+ Argument 'average' added to plot function 'simXyplot' that specifies whether the mean or median is used to compute averages. + Replaced example data, which is now called 'eusilcP' and has household income components. + Added accessor and mutator functions 'getFoo' and 'setFoo' for slots (here "foo" stands for the name of the slot). If no method 'setFoo' is available, this means that the slot is not supposed to be changed by the user. + Renamed slots "group" to "grouping" and "method" to "fun" (where available) since functions named 'getGroup', 'getMethod' and 'setMethod' already exist and cannot be used for accessor and mutator functions, respectively. + Renamed function 'getProb' to 'inclusionProb' to allow accessor method for slot "prob". + Removed class "SimResult" to avoid confusion. + Added 'show' methods and improved existing ones. + Added 'summary', 'head' and 'tail' methods. + Restructured help. + Added demos for design-based simulation, model-based simulation and parallel computing. + Classes "SampleControl" and "SampleSetup" now contain a slot that indicates whether groups are collected after sampling individuals or sampled directly. + Class "SampleSetup" now contains a slot that stores the seed of the random number generator before and after setting up the samples. + Made various improvements for plot functions if simulation results with multiple contamination levels or missing value rates are supplied. + Included package vignettes.
Changes in simFrame version 0.1.2
+ Class "NAControl" now contains a slot that specifies whether missing values should also be inserted into contaminated observations. + Added 'aggregate' method for objects of class "SimResults". + In function 'contaminate', the column ".contaminated" is always appended, even if 'epsilon' is 0 or 'x' does not have any rows.
Changes in simFrame version 0.1.1
+ Argument 'select' added to plot functions for specifying columns to be plotted. + Row names of example data 'eusilc' are now sorted from 1 to the number of rows. + Fixed cross-references in some help files.