Exercise

Using a DataFrame

In the previous exercise, you saw how to split up a task and use the low-level python multiprocessing.Pool API to do calculations on several processing units.

It's essential to understand this on a lower level, but in reality, you'll never use this kind of APIs. A more convenient way to parallelize an apply over several groups is using the dask framework and its abstraction of the pandas DataFrame, for example.

The pandas DataFrame, athlete_events, is available in your workspace.

Instructions 1/2

undefined XP

1

2

Create 4 partitions of the athletes_events DataFrame using dd.from_pandas().

If you forgot the parameters of dd.from_pandas(), check out the slides again, or type help(dd.from_pandas) in the console!