COMPSTAT 2016: Start Registration
View Submission - COMPSTAT
Title: FinSam: An R package for finite population sampling analysis Authors:  Jacinto Martin Jimenez - Universidad de Extremadura (Spain) [presenting]
Manuel Molina - University of Extremadura (Spain)
Juan Carlos Ridruejo Sayavera - De Extremadura (Spain)
Abstract: FinSam is an R package intended to support estimation analysis in finite population sampling. It covers the usual basic sampling schemes like simple random sampling and unequal probabilities, trough Horvitz-Thompson estimator. Besides, it includes cluster sampling (one and two stages) and indirect methods, ratio and regression estimators. The package also includes some complementary material, like inverse sampling, randomized responses, Jackknife method and others. It contains data examples and functions that allow us to obtain the different estimators and their variances in the command-line frontend. The final objective of the package is to use it in an advanced course on Finite Population Sampling. So, it includes a user-friendly menu-driven interface. It allows us to read data from files or introduce them manually according to the selected sampling analysis. Each method has a sub-menu with several options, mainly: Type of sampling, Managing and visualizing data and Computations. This is the first version of the FinSam package, and we are willing to expand it with other methods and examples.