arrow

R integration with Apache Arrow.

Installation

You first need to install the C++ library:

macOS

On macOS, you may install using homebrew:

brew install apache-arrow

From source

First install a release build of the C++ bindings to arrow.

git clone https://github.com/apache/arrow.git
cd arrow/cpp && mkdir release && cd release

# It is important to statically link to boost libraries
cmake .. -DARROW_PARQUET=ON -DCMAKE_BUILD_TYPE=Release -DARROW_BOOST_USE_SHARED:BOOL=Off
make install

Then the R package

From github (defaults to master branch)

library("devtools")
devtools::install_github("apache/arrow/r")

Example

library(arrow)
#> 
#> Attaching package: 'arrow'
#> The following object is masked from 'package:utils':
#> 
#>     timestamp
#> The following objects are masked from 'package:base':
#> 
#>     array, table

(tib <- tibble::tibble(x = 1:10, y = rnorm(10)))
#> # A tibble: 10 x 2
#>        x      y
#>    <int>  <dbl>
#>  1     1 -1.59 
#>  2     2  1.31 
#>  3     3 -0.513
#>  4     4  2.64 
#>  5     5  0.389
#>  6     6  0.660
#>  7     7  1.66 
#>  8     8 -0.120
#>  9     9  2.43 
#> 10    10  1.53
tab <- table(tib)
tab$schema
#> arrow::Schema 
#> x: int32
#> y: double
tab
#> arrow::Table
as_tibble(tab)
#> # A tibble: 10 x 2
#>        x      y
#>    <int>  <dbl>
#>  1     1 -1.59 
#>  2     2  1.31 
#>  3     3 -0.513
#>  4     4  2.64 
#>  5     5  0.389
#>  6     6  0.660
#>  7     7  1.66 
#>  8     8 -0.120
#>  9     9  2.43 
#> 10    10  1.53

Developing

The arrow package is using devtools for package related manipulations. To install:

library.packages("devtools")

Within a local clone, one can test changes by running the test suite.

Running the test suite

library("devtools")
devtools::install_dev_deps()
devtools::test()

Full package validation

R CMD build --keep-empty-dirs .
R CMD check arrow_*.tar.gz --as-cran --no-manual