Dask GPU Dataframes

A partitioned gpu-backed dataframe, using Dask.

Setup from source

Setup from source repo:

1. Install dependencies into a new conda environment where CUDA_VERSION is either 9.2 or 10

   conda create -n dask-cudf \
      -c rapidsai -c numba -c conda-forge -c defaults \
      cudf dask cudatoolkit={CUDA_VERSION}
   

2. Activate conda environment:

   source activate dask-cudf
   

3. Clone dask-cudf repo:

   git clone https://github.com/rapidsai/dask-cudf
   

4. Install from source:

   cd dask-cudf
   pip install .
   

Test

1. Install pytest

   conda install pytest
   

2. Run all tests:

   py.test dask_cudf
   

3. Or, run individual tests:

    py.test dask_cudf/tests/test_file.py
   

Style

For style we use black, isort, and flake8. These are available as pre-commit hooks that will run every time you are about to commit code.

From the root directory of this project run the following:

pip install pre-commit
pre-commit install