Julia bindings for the xtensor C++ multi-dimensional array library.

• xtensor is a C++ library for multi-dimensional arrays enabling numpy-style broadcasting and lazy computing.

• xtensor-julia enables inplace use of julia arrays in C++ with all the benefits from xtensor

  • C++ universal function and broadcasting
  • STL - compliant APIs.
  • A broad coverage of numpy APIs (see the numpy to xtensor cheat sheet).

The Julia bindings for xtensor are based on the libcxxwrap C++ library.

To get started with using xtensor-julia, check out the full documentation

http://xtensor-julia.readthedocs.io/

xtensor-julia offers two container types wrapping julia arrays inplace to provide an xtensor semantics

• jltensor
• jlarray.

Both containers enable the numpy-style APIs of xtensor (see the numpy to xtensor cheat sheet).

• On the one hand, jlarray has a dynamic number of dimensions. It can be reshaped dynamically and the new shape is reflected on the Julia side.

• On the other hand jltensor has a compile time number of dimensions, specified with a template parameter. Shapes of jltensor instances are stack allocated, making jltensor a significantly faster expression than jlarray.

C++ code

#include <numeric>                        // Standard library import for std::accumulate
#include <cxx_wrap.hpp>                   // libcxxwrap import to define Julia bindings
#include "xtensor-julia/jltensor.hpp"     // Import the jltensor container definition
#include "xtensor/xmath.hpp"              // xtensor import for the C++ universal functions

double sum_of_sines(xt::jltensor<double, 2> m)
{
    auto sines = xt::sin(m);  // sines does not actually hold values.
    return std::accumulate(sines.cbegin(), sines.cend(), 0.0);
}

JLCXX_MODULE define_julia_module(jlcxx::Module& mod)
{
    mod.method("sum_of_sines", sum_of_sines);
}

Julia Code

using xtensor_julia_test

arr = [[1.0 2.0]
       [3.0 4.0]]

s = sum_of_sines(arr)
s

Outputs

1.1350859243855171

C++ code

#include <cxx_wrap.hpp>
#include "xtensor-julia/jlvectorize.hpp"

double scalar_func(double i, double j)
{
    return std::sin(i) - std::cos(j);
}

JLCXX_MODULE define_julia_module(jlcxx::Module& mod)
{
    mod.method("vectorized_func", xt::jlvectorize(scalar_func));
}

Julia Code

using xtensor_julia_test

x = [[ 0.0  1.0  2.0  3.0  4.0]
     [ 5.0  6.0  7.0  8.0  9.0]
     [10.0 11.0 12.0 13.0 14.0]]
y = [1.0, 2.0, 3.0, 4.0, 5.0]
z = vectorized_func(x, y)
z

Outputs

[[-0.540302  1.257618  1.89929   0.794764 -1.040465],
 [-1.499227  0.136731  1.646979  1.643002  0.128456],
 [-1.084323 -0.583843  0.45342   1.073811  0.706945]]

xtensor-julia a header-only C++ library. It has been packaged for the mamba (or conda) package manager.

mamba install xtensor-julia -c conda-forge

xtensor-r can be installed from source with cmake in any installation prefix. For example, on unix systems

cmake -D CMAKE_INSTALL_PREFIX=/prefix/path/
make
make install

We also provide a Julia package for xtensor, which has been packaged for both conda and Pkg (Julia's package manager). The repository for the Julia package is https://github.com/xtensor-stack/Xtensor.jl.

To install the Julia package:

using Pkg; Pkg.add("Xtensor");

The Julia available from the Julia package manager vendors the headers for xtensor-julia, xtensor, xtlandxsimd`.

xtensor-julia's documentation is built with three tools

• doxygen
• sphinx
• breathe

While doxygen must be installed separately, you can install breathe by typing

pip install breathe

Breathe can also be installed with conda

conda install -c conda-forge breathe

Finally, build the documentation with

make html

from the docs subdirectory.

cmake -D BUILD_TESTS=ON ..

xtensor-julia depends on the xtensor and libcxxwrap-julia libraries

These dependencies are automatically resolved when using the Julia package manager.

We use a shared copyright model that enables all contributors to maintain the copyright on their contributions.

This software is licensed under the BSD-3-Clause license. See the LICENSE file for details.