cuGraph Python strives to make getting data into and out of cuGraph simple. To that end, the Python interface accepts
cuGraph supports graph creation with Source and Destination being expressed as:
- cuDF DataFrame
- Pandas DataFrame
- NetworkX graph classes
- Numpy arrays
- CuPy sparse matrix
- SciPy sparse matrix
cuGraph tries to match the return type based on the input type. So a NetworkX input will return the same data type that NetworkX would have.
The preferred data type is a cuDF object since it is already in the GPU. For loading data from disk into cuDF please see the cuDF documentation.
Loading data
- Graph.from_cudf_adjlist
- Graph.from_cudf_edgelist
Results
Results which are not simple types (ints, floats) are typically cuDF Dataframes.
The RAPIDS cuDF library can be thought of as accelerated Pandas