Use Of R in Statistics Lithuania

Tomas Rudys (tomas.rudys@stat.gov.lt)
Statistics Lithuania

Abstract

Recently R becoming more and more popular among official statistics offices. It can be used not even for research purposes, but also for a production of official statistics. Statistics Lithuania recently started an analysis of possibilities where R can be used and could it replace some other statistical programming languages or systems. For this reason a work group was arranged. In the paper we will present overview of the current situation on implementation of R in Statistics Lithuania, some problems we are chasing with and some future plans. At the current situation R is used mainly for research purposes. Looking for- ward a short courses on basic R was prepared and at the moment we are starting to use R for data analysis, data manipulation from Oracle data bases, some reports preparation, data editing, survey estimation. On the other hand we found some problems working with big data sets, also survey sampling as there are surveys with complex sampling designs. We are also analysing the running of R on our servers in order to have possibilities to use more random access memory (RAM). Despite the problems, we are trying to use R in more fields in production of official statistics.

Keywords: R, official statistics, sampling, outlier detection
JEL Classification: C10, C88

[Full Text]

Romanian Statistical Review 2/2016