Skip to main content
Join the official 2019 Python Developers SurveyStart the survey!

cuDF - GPU Dataframe

Project description

 cuDF - GPU DataFrames

Build Status  Documentation Status

The RAPIDS cuDF library is a GPU DataFrame manipulation library based on Apache Arrow that accelerates loading, filtering, and manipulation of data for model training data preparation. The RAPIDS GPU DataFrame provides a pandas-like API that will be familiar to data scientists, so they can now build GPU-accelerated workflows more easily.

NOTE: For the latest stable ensure you are on the master branch.

Quick Start

Please see the Demo Docker Repository, choosing a tag based on the NVIDIA CUDA version you’re running. This provides a ready to run Docker container with example notebooks and data, showcasing how you can utilize cuDF.

Install cuDF


It is easy to install cuDF using conda. You can get a minimal conda installation with Miniconda or get the full installation with Anaconda.

Install and update cuDF using the conda command:

# CUDA 9.2
conda install -c nvidia -c rapidsai -c numba -c conda-forge -c defaults cudf

# CUDA 10.0
conda install -c nvidia/label/cuda10.0 -c rapidsai/label/cuda10.0 -c numba -c conda-forge -c defaults cudf

Note: This conda installation only applies to Linux and Python versions 3.6/3.7.


It is easy to install cuDF using pip. You must specify the CUDA version to ensure you install the right package.

# CUDA 9.2
pip install cudf-cuda92

# CUDA 10.0.
pip install cudf-cuda100

Development Setup

The following instructions are for developers and contributors to cuDF OSS development. These instructions are tested on Linux Ubuntu 16.04 & 18.04. Use these instructions to build cuDF from source and contribute to its development. Other operatings systems may be compatible, but are not currently tested.

Get libcudf Dependencies

Compiler requirements:

  • gcc version 5.4+
  • nvcc version 9.2+
  • cmake version 3.12.4+

CUDA/GPU requirements:

  • CUDA 9.2+
  • NVIDIA driver 396.44+
  • Pascal architecture or better

Python requirements:

  • 3.6 or 3.7

You can obtain CUDA from

Since cmake will download and build Apache Arrow you may need to install Boost C++ (version 1.58+) before running cmake:

# Install Boost C++ for Ubuntu 16.04/18.04
$ sudo apt-get install libboost-all-dev


# Install Boost C++ for Conda
$ conda install -c conda-forge boost

Script to build cuDF from source

Build from Source

To install cuDF from source, ensure the dependencies are met and follow the steps below:

  • Clone the repository and submodules
git clone $CUDF_HOME
git submodule update --init --remote --recursive
  • Create the conda development environment cudf_dev
# create the conda environment (assuming in base `cudf` directory)
conda env create --name cudf_dev --file conda/environments/cudf_dev_cuda9.2.yml # for CUDA 9.2
# or
conda env create --name cudf_dev --file conda/environments/cudf_dev_cuda10.0.yml # for CUDA 10.0
# activate the environment
source activate cudf_dev
  • Build and install libcudf. CMake depends on the nvcc executable being on your path or defined in $CUDACXX.
$ cd $CUDF_HOME/cpp                                                       # navigate to C/C++ CUDA source root directory
$ mkdir build                                                             # make a build directory
$ cd build                                                                # enter the build directory

# CMake options:
# -DCMAKE_INSTALL_PREFIX set to the install path for your libraries or $CONDA_PREFIX if you're using Anaconda, i.e. -DCMAKE_INSTALL_PREFIX=/install/path or -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX
# -DCMAKE_CXX11_ABI set to ON or OFF depending on the ABI version you want, defaults to ON. When turned ON, ABI compability for C++11 is used. When OFF, pre-C++11 ABI compability is used.
$ cmake .. -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX -DCMAKE_CXX11_ABI=ON      # configure cmake ...

$ make -j                                                                 # compile the libraries, ... '-j' will start a parallel job using the number of physical cores available on your system
$ make install                                                            # install the libraries, to the CMAKE_INSTALL_PREFIX
  • To run tests (Optional):
$ make test
  • Build, install, and test cffi bindings:
$ make python_cffi                                  # build CFFI bindings for,
$ make install_python                               # build & install CFFI python bindings. Depends on cffi package from PyPi or Conda
$ cd python && py.test -v                           # optional, run python tests on low-level python bindings
  • Build the cudf python package, in the python folder:
$ cd $CUDF_HOME/python
$ python build_ext --inplace
  • You will also need the following environment variables, including $CUDA_HOME.
  • To run Python tests (Optional):
$ py.test -v                                        # run python tests on cudf python bindings
  • Finally, install the Python package to your Python path:
$ python install                           # install cudf python bindings

Done! You are ready to develop for the cuDF OSS project.

Debugging cuDF

Building Debug mode from source

Follow the above instructions to build from source and add -DCMAKE_BUILD_TYPE=Debug to the cmake step.

For example:

$ cmake .. -DCMAKE_INSTALL_PREFIX=/install/path -DCMAKE_BUILD_TYPE=Debug     # configure cmake ... use -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX if you're using Anaconda

This builds libcudf in Debug mode which enables some assert safety checks and includes symbols in the library for debugging.

All other steps for installing libcudf into your environment are the same.

Debugging with cuda-gdb and cuda-memcheck

When you have a debug build of libcudf installed, debugging with the cuda-gdb and cuda-memcheck is easy.

If you are debugging a Python script, simply run the following:


cuda-gdb -ex r --args python <program_name>.py <program_arguments>


cuda-memcheck python <program_name>.py <program_arguments>

Automated Build in Docker Container

A Dockerfile is provided with a preconfigured conda environment for building and installing cuDF from source based off of the master branch.


  • Install nvidia-docker2 for Docker + GPU support
  • Verify NVIDIA driver is 396.44 or higher
  • Ensure CUDA 9.2+ is installed


From cudf project root run the following, to build with defaults:

$ docker build --tag cudf .

After the container is built run the container:

$ docker run --runtime=nvidia -it cudf bash

Activate the conda environment cudf to use the newly built cuDF and libcudf libraries:

root@3f689ba9c842:/# source activate cudf
(cudf) root@3f689ba9c842:/# python -c "import cudf"
(cudf) root@3f689ba9c842:/#

Customizing the Build

Several build arguments are available to customize the build process of the container. These are specified by using the Docker build-arg flag. Below is a list of the available arguments and their purpose:

Build Argument Default Value Other Value(s) Purpose
CUDA_VERSION 9.2 10.0 set CUDA version
LINUX_VERSION ubuntu16.04 ubuntu18.04 set Ubuntu version
CC & CXX 5 7 set gcc/g++ version; NOTE: gcc7 requires Ubuntu 18.04
CUDF_REPO This repo Forks of cuDF set git URL to use for git clone
CUDF_BRANCH master Any branch name set git branch to checkout of CUDF_REPO
NUMBA_VERSION newest >=0.40.0 set numba version
NUMPY_VERSION newest >=1.14.3 set numpy version
PANDAS_VERSION newest >=0.23.4 set pandas version
PYARROW_VERSION 0.12.0 Not supported set pyarrow version
CMAKE_VERSION newest >=3.12 set cmake version
CYTHON_VERSION 0.29 Not supported set Cython version
PYTHON_VERSION 3.6 3.7 set python version

Open GPU Data Science

The RAPIDS suite of open source software libraries aim to enable execution of end-to-end data science and analytics pipelines entirely on GPUs. It relies on NVIDIA® CUDA® primitives for low-level compute optimization, but exposing that GPU parallelism and high-bandwidth memory speed through user-friendly Python interfaces.

Apache Arrow on GPU

The GPU version of Apache Arrow is a common API that enables efficient interchange of tabular data between processes running on the GPU. End-to-end computation on the GPU avoids unnecessary copying and converting of data off the GPU, reducing compute time and cost for high-performance analytics common in artificial intelligence workloads. As the name implies, cuDF uses the Apache Arrow columnar data format on the GPU. Currently, a subset of the features in Apache Arrow are supported.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for cudf, version 0.6.1
Filename, size File type Python version Upload date Hashes
Filename, size cudf-0.6.1-cp36-cp36m-manylinux1_x86_64.whl (17.2 MB) File type Wheel Python version cp36 Upload date Hashes View hashes
Filename, size cudf-0.6.1-cp37-cp37m-manylinux1_x86_64.whl (17.2 MB) File type Wheel Python version cp37 Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page