Dataset Cleanup and Comepletion

[BradReesWork]

This issues in an extension to the current Datasets Epic - #3357

Goals: at the Python (aka cugraph) layer

• All Python code should be using the Datasets API

• All example code should be using the Datasets API

• All functions should be well documented

• All functions should be fully coded and not be stubs

• Provide results as a Graph: cugraph.Graph or nx.Graph

• Provide results as a dataframe: cuDF or Pandas

• Provide option to return data as a SciPy Sparse Matrix

• Data file downloaded if not found in the specified dataset folder

• Data file can be cleaned up at the end of use (function argument)

• The "dataset" folder currently consumes 4.6 GB , would be nice to reduce that

• Remove all file currently in Datasets

• move all other files to an "extra" folder and start determining if they are used

• validate download script for the C code

• fix race condition