Abstract:
|
In this paper, we introduce the R package, RPMS, (Recursive Partitioning for Modeling Survey Data), which fits a linear model to survey data conditionally on variables selected through recursively partitioning. This application of recursive partitioning produces design consistent regression tree models of the data. We demonstrate the software through the modeling of establishment level employment data collected by the U.S. Bureau of Labor Statistics. These applications demonstrate the easily interpretable structure and the ability of regression trees to handle large data sets with complex associations among the variables.
|