RaProR - Random Projections for Bayesian linear Regression

[Members] [Support] [R-Package] [Literature]


  • Dipl.-Stat. Leo Geppert
  • Prof. Dr. Katja Ickstadt
  • Dipl.-Inf. Alexander Munteanu
  • B.Sc.-Inf. Jens Quedenfeld
  • Prof. Dr. Christian Sohler


Research supported by Deutsche Forschungsgemeinschaft, SFB 876, project C4.


The R-package RaProR can be used to calculate a sketch of a large data set. That is a substitute data set of the same dimension but smaller number of observations. As we show in [1], the sketch can be used to perform approximate Bayesian or frequentist linear regression. More specifically, the likelihood as well as the posterior will be close to the ones obtained on the original data. Any algorithm for the regression analysis will run much faster on the sketch than on the original data set given that its running time depends on the number of observations.



The pre-compiled Windows packages can be installed by either using install.packages or by choosing install package(s) from local files... from the packages-menu in the R GUI. Note that the downloaded archives must not be renamed. For compiling the source tarball, ensure that you have at least GCC 4.7 installed and all the features and STL of C++ 11 enabled.


[1] Geppert, LN, Ickstadt, K., Munteanu, A., Quedenfeld, J., Sohler, C.: Random projections for Bayesian regression, In Statistics and Computing, DOI 10.1007/s11222-015-9608-z, Springer 2015. (Arxiv version: stat.CO:abs/1504.06122)

Last update Jan 21, 2016 by A Munteanu and LN Geppert