Statistical software and database systems
Statistics software and relational database systems possess considerable overlap in the areas of data loading and processing. While statistical packages are optimized for advanced algorithms, database systems provide fast access to large volumes of data. Taking the best of both worlds, this combination leads to a powerful symbiotic platform to speedup data discovery. For this purpose, MonetDB provides two-fold integration with the R statistical software:
The R integration module allows for running embedded R queries in MonetDB. The module works by wrapping R code in a SQL user-defined function. Since we no longer need to transfer data between the two systems, we gain great performance benefits. Using the MonetDB/R module we achieve a zero-copy integration between MonetDB and R, while bringing the power of statistical software to our high-performance database. The user is freed from the tedious task of shuffling data around. This means that one spends more time evaluating the result of their queries and not waiting for them to finish.
Check out the MonetDB.org blog post with more details.
MonetDB R integration (Rint) performance compared to other systems.
The MonetDB.R package provides a transparent connection to MonetDB from the R software suite. The connector achieves great performance by understanding data processing in R. It decides which portions of the data analysis should be performed by either R or the MonetDB database. In addition to running queries on data stored in MonetDB, the MonetDB.R connector also supports the higher-level dplyr data manipulation package. Through this integration, each system is used to its best performance. This also reduces the overhead of shuffling data between different systems, improving data processing time. R users will feel right at home, noticing only the improved speed when analyzing large volumes of data.
See why MonetDB.R users consider that "MonetDB analysis speeds will outrun even the fastest of imaginary crime-fighting superheroes".
The MonetDB/R module is available as of the October 2014 release of MonetDB.